U.S. Religious Landscape Survey Analysis

knitr::opts_chunk$set(echo = TRUE)

“The Last Judgment” by Michelangelo

Abstract

Alexis de Tocqueville, a prominent French writer from the early 19th century who extensively studied democracy in America, recognized the significant influence religion has on American society. He observed that American civilization uniquely combines the spirit of religion and the spirit of liberty, blending them harmoniously.

Nearly two centuries later, religion continues to shape the beliefs and values of Americans. To explore the relationship between religious beliefs and social and political views, the Pew Research Center’s Forum on Religion & Public Life conducted a comprehensive survey on religion in America. The first report, published in February 2008, examined the religious affiliations of the American public, as well as changes in affiliation and the diversity of religious traditions.

The latest release of the Landscape Survey provides detailed information on various aspects of religious beliefs and practices in America. This includes the importance of religion in people’s lives, belief in God and the afterlife, worship attendance, and participation in religious activities. The study also investigates public opinion on religion’s impact on society, conflicts between religion and society, and the relationship between religion and life satisfaction.

Using survey responses, the report analyzes the diversity of opinion on political and public policy issues among different religious groups in the United States, including those who are unaffiliated or nominally affiliated with a religious tradition. These issues range from ideological orientation to views on social issues, government size and role, environmental concerns, and foreign affairs.

The Landscape Survey is based on a nationwide survey conducted from May 8 to August 13, 2007, with over 35,000 adult participants. Additional oversampling was conducted for Eastern Orthodox Christians, Buddhists, and Hindus. The study also incorporates data from a 2007 survey of American Muslims, conducted in partnership with other Pew Research Center projects, resulting in interviews with over 36,000 Americans in total.

Keywords

1. Religion in America
2. American civilization
3. Religious beliefs and practices
4. Social and political views
5. Diversity of religious traditions
6. Public opinion on religious impact
7. Political and public policy issues
8. Statistical modeling and analysis

Introduction

I would like to divide this project into several chapters, in each chapter I’m gonna search for the answer of my own questions.

Chapter 1 (Religious Beliefs and Practices)

Question: How do Americans’ religious beliefs and practices vary across different demographic groups, and what insights can be gained by examining measures such as the importance of religion in people’s lives, belief in God, frequency of prayer, and frequency of worship service attendance?

Explanation: This chapter delves into the diverse religious beliefs and practices of American adults. It begins by exploring the importance individuals attribute to religion in their lives and delves into their perspectives on God, Scripture, miracles, and other religious beliefs. Furthermore, it discusses worship and other congregational activities, as well as devotional practices and spiritual experiences. The chapter also addresses beliefs about religion, including exclusivity in claims to salvation and moral considerations. Along the way, four key measures of religious commitment – importance of religion in people’s lives, belief in God, frequency of prayer, and frequency of worship service attendance – are highlighted for demographic analysis. These measures serve as lenses through which to examine social and political attitudes within religious traditions in subsequent chapters.

Chapter 2 (Social and Political Views)

Question: How do Americans’ religious affiliations and levels of religious commitment influence their views on social and political issues?

Explanation: This chapter delves into the relationship between Americans’ religious affiliations, beliefs, practices, and their social and political attitudes. Despite religion being a secondary source for many, the analysis reveals strong correlations between religious affiliation, beliefs, and social attitudes.

Chapter 3 (The Changing Religious Trends)

Question: What trends have emerged in the religious landscape of the United States in recent years, and how have they impacted the overall attitude of the population toward different subjects?

Explanation: This chapter examines the changing religious landscape and social attitudes in the United States over the past seven years. It explores shifts in religious affiliation, attitudes towards abortion, views on homosexuality, acceptance of evolution, and experiences of spiritual peace. With Christians declining in proportion and the rise of the religiously unaffiliated, societal attitudes have evolved, reflecting greater acceptance and diversity of beliefs. Through this analysis, we gain insight into the evolving dynamics of religion, morality, and personal values in American society.

Chapter 4 (Religious Switching and Intermarriage)

Question: How can we model and understand the trends and dynamics of religious affiliation changes among American adults?

Explanation: In this revised chapter, our analysis shifts from a broad discussion of religious switching and intermarriage to a more focused examination of the factors influencing religious affiliation changes over time. Utilizing advanced statistical models, including decision trees and Generalized Linear Models (GLM), we delve into the intricacies of religious identity transitions from childhood to adulthood. By examining the predictive power of sociodemographic variables such as income, education, age, and marital status, our models shed light on the underlying patterns that guide religious affiliation changes. We explore the significance of these variables in contributing to an individual’s likelihood of shifting religious identity, providing a nuanced understanding of the factors at play.

Data Description

2007 Religious Landscape Study

Summary: The 2007 U.S. Religious Landscape Study is a comprehensive telephone survey that explores the religious affiliations, beliefs, and practices of the American public. Conducted by the Pew Research Center, this survey provides insights into the religious composition of the U.S. and examines how religion intersects with various aspects of public life.

Dataset Details:

Aspect	Details
Survey Conducted By	Pew Research Center
Date of Survey	2007
Sample Size	Over 35,000 U.S. Adults
Methodology	Telephone survey (landline and cellphones)
Geographic Coverage	National (U.S.)
Population Coverage	97% of non-institutionalized U.S. adult population
Non-Coverage	3% of U.S. adults not reachable by telephone or non-English/Spanish speakers
Data Availability	Publicly available for analysis
Funding	The Pew Charitable Trusts, supported by Lilly Endowment Inc.
Research Purpose	To inform about religious affiliations, beliefs, and practices; and understand their influence on various aspects of public life

2014 Religious Landscape Study

Summary: The 2014 Religious Landscape Study, following up on the initial 2007 survey, provides updated data on religious affiliations and the changing religious composition in the United States. This extensive survey captures current trends and shifts in religion within the American adult population.

Dataset Details:

Aspect	Details
Survey Conducted By	Pew Research Center
Date of Survey	June 4 - Sept. 30, 2014
Sample Size	35,071 U.S. Adults
Methodology	Telephone survey (60% cellphones, 40% landlines)
Geographic Coverage	National (U.S.), with at least 300 interviews in every state and the District of Columbia
Language	English and Spanish
Population Coverage	97% of non-institutionalized U.S. adult population
Non-Coverage	3% of U.S. adults not reachable by telephone or non-English/Spanish speakers
Data Availability	Publicly available for analysis
Funding	The Pew Charitable Trusts, supported by Lilly Endowment Inc.
Research Purpose	To provide insights into the religious beliefs and practices, as well as the social and political attitudes of the U.S. adults

These datasets are instrumental for analyzing changes over time in religious beliefs and practices among U.S. adults. They offer a wealth of information that can be utilized to study the relationship between religion and a variety of sociodemographic factors. The data is publicly available and provides a representative snapshot of the religious composition of the United States for the respective years.

Chapter 1 : Religious Beliefs and Practices

Religion Composition

This statistical breakdown provides a clear picture of the religious affiliations among survey respondents, focusing on those with more than 100 individuals per religion. It highlights the diversity of religious identities represented in the survey, including Agnostic, Atheist, Buddhist, Hindu, Jewish, Mormon, Muslim, Orthodox, Protestant, Roman Catholic, and others. The larger representation of categories like Protestant and Roman Catholic indicates their significant presence in the surveyed population. Additionally, the inclusion of categories such as “Nothing in particular”, “Agnostic”, “Atheist” underscores the prevalence of religious non-affiliation among respondents. Overall, this breakdown offers a straightforward overview of the religious diversity within the survey sample, facilitating further analysis of attitudes and beliefs across different religious affiliations.

# Load the necessary library
library(readr)
library(igraph)
library(ggraph)
library(colormap)
library(oce)
library(pROC)
library(ggplot2)
library(magrittr) # needs to be run every time you start R and want to use %>%
library(dplyr)    # alternatively, this also loads %>%
library(plotly)
library(kableExtra)
library(tidyr)
library(rpart)
library(randomForest)
library(caret)

# Load the dataset 2007
survey_data_2007 <- read_csv("Dataset_2007.csv")
# Load the dataset 2014
survey_data_2014 <- read_csv("Dataset_2014.csv")

# Let's say we have the data frame 'survey_data'
survey_data <- data.frame(survey_data_2007$q16)


# Assuming 'q16' is the column for religion
religion_counts_2007 <- survey_data_2007 %>%
  count(q16) %>%
  filter(n > 100)

# Rename the columns for clarity
colnames(religion_counts_2007) <- c("Religion", "Respondents")

# Calculate the total number of respondents
total_respondents <- sum(religion_counts_2007$Respondents)

# Calculate percentages
religion_counts_2007 <- religion_counts_2007 %>%
  mutate(Percentage = (Respondents / total_respondents) * 100)

options(repr.plot.width = 8, repr.plot.height =5)
# Create the bar plot
p <- ggplot(religion_counts_2007, aes(x = reorder(Religion, -Percentage), y = Percentage, fill = Religion) , width = 60 , height = 12) + 
  geom_bar(stat = "identity") + # Use identity to use the heights of the bars as provided in the data frame
  coord_flip() +  # Flip coordinates to have horizontal bars
  theme_minimal() +  # Use a minimal theme for aesthetics
  labs(title = "Number of Respondents",
       x = "",
       y = "") +
  theme(legend.position = "none")  # Remove the legend


# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 500)

# Show the interactive plot in your default browser

interactive_plot

Importance of Religion

Question:

“How important is religion in your life?”

• Very Important: A significant proportion of respondents from various religious backgrounds, including Jehovah’s Witnesses, Mormons, Muslims, Orthodox Christians, Protestants, Roman Catholics, and Unitarians, consider religion to be highly important in their lives. These individuals prioritize religious beliefs and practices as integral aspects of their identity and worldview.

• Somewhat Important: Another sizable segment of respondents, spanning diverse religious groups such as Buddhists, Hindus, and those with no particular religious affiliation, view religion as somewhat important in their lives. While not as central as for those who consider it very important, these individuals still recognize the value of religious beliefs and traditions to some extent.

• Not Too Important or Not at All Important: A minority of respondents, particularly atheists and agnostics, express that religion holds little to no importance in their lives. They either do not believe in religious concepts or do not actively engage with religious practices, indicating a lack of significance of religion in shaping their personal identities or guiding their daily lives.

religion_importance <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q21)) %>%  # Ensure q16 and q21 are not NA
  group_by(q16, q21) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q21)



p <- ggplot(religion_importance, aes(x = q16, y = q21, fill = Percentage)) +
  geom_tile() + # Create the tiles for the heatmap
  scale_fill_gradient(low = "white", high = "steelblue") + # Gradient from white (low) to steel blue (high)
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1), axis.text.y = element_text(size = 12), title = element_text(size = 20)) +
  labs(title = "Percentage of Responses on the Importance of Religion by Religious Group",
       x = "Religion",
       y = "Importance",
       fill = "Percentage")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 750)

# Show the interactive plot in your default browser

interactive_plot

Conclusion:

The data underscores the diversity of attitudes toward religion among individuals of various religious affiliations. While some prioritize religion as a fundamental aspect of their lives, others assign varying degrees of importance to it, and some do not consider it significant at all. Understanding these perspectives is essential for appreciating the complex role that religion plays in shaping individual identities and societal dynamics.

religion_importance_selected <- select(religion_importance, q16, q21, Percentage)


religion_importance_wide <- pivot_wider(
  religion_importance_selected,
  names_from = q21,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(religion_importance_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(religion_importance_wide, "html", align = c('l', rep('c', 10))) %>%
  kable_styling("striped", full_width = F) %>%
  add_header_above(c(" " = 1, "How important is religion in your life" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	How important is religion in your life
Religion	Don’t know/Refused (VOL.)	Not at all important	Not too important	Somewhat important	Very important
Agnostic (not sure if there is a God)	1.4527845	43.8256659	33.414044	16.101695	5.205811
Atheist (do not believe in God)	2.3300971	74.5631068	11.844660	7.766990	3.495146
Buddhist	2.1897810	6.3260341	17.761557	37.469586	36.253041
Christian (VOL)	3.2534247	4.7945205	7.876712	24.143836	59.931507
Don’t know/Refused (VOL)	19.2000000	18.8000000	15.600000	25.600000	20.800000
Hindu	0.7782101	3.1128405	12.451362	41.634241	42.023346
Jehovah’s Witness (VOL)	0.9389671	0.4694836	1.408451	8.450704	88.732394
Jewish (Judaism)	0.8797654	9.8240469	18.035191	40.469208	30.791789
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	0.0000000	1.0327022	2.409639	10.843374	85.714286
Muslim (Islam)	0.0000000	4.3103448	7.758621	23.275862	64.655172
Nothing in particular	1.7534394	28.2168870	24.143512	27.596439	18.289722
Orthodox (Greek, Russian, or some other orthodox church)	0.0000000	3.3057851	8.815427	30.578512	57.300275
Protestant	0.5557674	1.3676238	4.724023	22.497684	70.854901
Roman Catholic	0.4970797	1.6527899	7.033677	32.757549	58.058904
Unitarian (Universalist) (VOL)	3.2467532	11.0389610	13.636364	37.662338	34.415584

Importance of Religion and Demographic Groups

The survey highlights a significant gender disparity regarding the importance of religion in individuals’ lives, with women more likely than men to consider religion very important. This trend persists across various religious groups.

Furthermore, age plays a substantial role in determining the importance of religion, with older adults more likely than younger adults to prioritize it. For instance, less than half (45%) of adults under 30 view religion as very important, contrasting with over two-thirds (68%) of those aged 60 and older.

Regarding education, the survey indicates that individuals with lower levels of education tend to place greater importance on religion in their lives.

religion_importance_gender <- survey_data_2007 %>%
  group_by(sex) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(sex), !is.na(q21)) %>%  # Ensure q16 and q21 are not NA
  group_by(sex, q21) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(sex) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(sex, q21)


religion_importance_gender_selected <- select(religion_importance_gender, sex, q21, Percentage)


religion_importance_gender_wide <- pivot_wider(
  religion_importance_gender_selected,
  names_from = q21,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(religion_importance_gender_wide)[1] <- "Sex"

# Create the table with kable and add styling with kableExtra
table <- kable(religion_importance_gender_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "How important is religion in your life" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	How important is religion in your life
Sex	Don’t know/Refused (VOL.)	Not at all important	Not too important	Somewhat important	Very important
Female	0.8928108	4.988321	6.405398	23.01064	64.70283
Male	1.0005525	9.250506	11.540114	28.59861	49.61021

filtered_data_age <- survey_data_2007 %>%
  filter(!is.na(age) & as.numeric(age) >= 10)  # This removes NA values and ensures age is 10 or more

## Warning: There was 1 warning in `filter()`.
## ℹ In argument: `!is.na(age) & as.numeric(age) >= 10`.
## Caused by warning:
## ! NAs introduced by coercion

filtered_data_age <- filtered_data_age %>%
  mutate(age_numeric = as.numeric(as.character(age)),  # Convert age from string to numeric
         AgeGroup = cut(age_numeric,
                        breaks = c(10, 20, 30, 40, 50, 60, 70, 80, Inf),
                        labels = c("10-20", "20-30", "30-40", "40-50", "50-60", "60-70", "70-80", "80+"),
                        right = FALSE)) %>%
  filter(!is.na(AgeGroup))  # Exclude entries that could not be converted to numeric (including "Don’t know/Refused (VOL.)")


religion_importance_age <- filtered_data_age %>%
  group_by(AgeGroup) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(AgeGroup), !is.na(q21)) %>%  # Ensure q16 and q21 are not NA
  group_by(AgeGroup, q21) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(AgeGroup) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(AgeGroup, q21)


religion_importance_age_selected <- select(religion_importance_age, AgeGroup, q21, Percentage)


religion_importance_age_wide <- pivot_wider(
  religion_importance_age_selected,
  names_from = q21,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(religion_importance_gender_wide)[1] <- "Age"

# Create the table with kable and add styling with kableExtra
table <- kable(religion_importance_age_wide, "html") %>%
  kable_styling("striped", full_width = F) %>%
  add_header_above(c(" " = 1, "How important is religion in your life" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box(height  = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	How important is religion in your life
AgeGroup	Don’t know/Refused (VOL.)	Not at all important	Not too important	Somewhat important	Very important
10-20	0.1042753	10.636079	12.200209	35.03650	42.02294
20-30	0.5787390	8.559245	12.671337	31.98294	46.20774
30-40	0.6489676	7.649951	10.383481	29.61652	51.70108
40-50	0.7569012	6.144257	9.349956	27.17424	56.57465
50-60	0.8131183	7.074129	9.147581	25.65388	57.31129
60-70	1.0150444	6.978430	7.123437	22.40348	62.47961
70-80	1.0983123	5.973748	5.679079	18.53737	68.71149
80+	1.2072435	4.778672	4.728370	17.85714	71.42857

religion_importance_edu <- survey_data_2007 %>%
  group_by(educ) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(educ), !is.na(q31)) %>%  # Ensure q16 and q31 are not NA
  group_by(educ, q31) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(educ) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(educ, q31)


religion_importance_edu_selected <- select(religion_importance_edu, educ, q31, Percentage)


religion_importance_edu_wide <- pivot_wider(
  religion_importance_edu_selected,
  names_from = q31,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(religion_importance_edu_wide)[1] <- "Education"

# Create the table with kable and add styling with kableExtra
table <- kable(religion_importance_edu_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "How important is religion in your life" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	How important is religion in your life
Education	Absolutely certain	Don’t know/refused (VOL.)	Fairly certain	Not at all certain	Not too certain
College graduate (B.S., B.A., or other 4-year degree)	74.93427	0.5550687	20.17236	1.2269939	3.111306
Don’t know/Refused (VOL.)	76.85185	2.7777778	18.98148	0.0000000	1.388889
High school graduate (Grade 12 or GED certificate)	81.14176	0.5879838	15.45863	0.5772931	2.234338
High school incomplete (Grades 9-11)	80.89725	0.8200675	15.43657	0.3376749	2.508442
None, or grade 1-8	75.63218	1.9540230	17.93103	1.3793103	3.103448
Post-graduate training or professional schooling after colle	70.87993	0.9706689	22.36759	1.2449884	4.536822
Some college, no 4-year degree (including associate degree)	80.16148	0.6920415	16.18608	0.6535948	2.306805
Technical, trade, or vocational school AFTER high school	82.21344	0.7905138	14.72332	0.6916996	1.581028

Religious Beliefs

God

The survey reveals a widespread belief in God or a universal spirit among American adults, with more than nine-in-ten (92%) affirming this belief. This includes the vast majority of individuals from every major religious tradition, with virtually all Mormon respondents and at least 95% of Catholics, Orthodox Christians, Jehovah’s Witnesses, and members of all three Protestant traditions expressing belief. Even among those not affiliated with a specific religious group, seven-in-ten acknowledge their belief in God or a universal spirit. Notably, a substantial portion of self-described atheists (21%) and more than half of self-described agnostics (55%) also profess belief in God or a universal spirit.

God_belief <- survey_data_2007 %>%
  filter(q30 == 'Yes') %>% 
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q30)) %>%  # Ensure q16 and q21 are not NA
  group_by(q16, q31) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q31)



p <- ggplot(God_belief, aes(x = q16, y = q31, fill = Percentage)) +
  geom_tile() + # Create the tiles for the heatmap
  scale_fill_gradient(low = "white", high = "steelblue") + # Gradient from white (low) to steel blue (high)
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1), axis.text.y = element_text(size = 12) , title = element_text(size = 15)) +
  labs(title = "believe in God or a universal spirit",
       x = "Religion",
       y = "",
       fill = "Percentage")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 750)

# Show the interactive plot in your default browser

interactive_plot

There is notable variability in the certainty and nature of individuals’ beliefs in God. While at least nine-in-ten Mormons, Jehovah’s Witnesses and Protestant express absolute certainty in the existence of God, smaller majorities of other religious groups also share this conviction. Specifically, Muslims (82%), Catholics (72%), Orthodox Christians (71%), and Hindus (57%) report being completely certain of the existence of God or a universal spirit. In contrast, fewer than half of Jews (41%) and Buddhists (39%) express absolute certainty in the existence of God or a universal spirit. Notably, approximately one-third of the unaffiliated (36%) assert absolute certainty in the existence of God or a universal spirit, with 8% of atheists claiming absolute certainty in this belief.

Conception of God

• God as a Person: The majority of respondents across different religious affiliations perceive God as a person. This belief is particularly strong among Jehovah’s Witnesses (86.06%), Mormons (92.39%), and Protestants (73.98%).

• God as an Impersonal Force: A significant portion of respondents, though smaller compared to those who view God as a person, conceive of God as an impersonal force. This perspective is notably prevalent among Buddhists (61.84%), Hindus (59.75%), and Unitarians (61.11%).

• Other/Both/Neither: Some respondents express conceptions of God that do not fit neatly into the categories of “person” or “impersonal force.” This group includes individuals who may perceive God in multifaceted ways or hold beliefs outside the typical paradigms of personal or impersonal deities.

God_conception <- survey_data_2007 %>%
  filter(q30 == 'Yes') %>% 
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q32)) %>%  # Ensure q16 and q21 are not NA
  group_by(q16, q32) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q32)


God_conception_selected <- select(God_conception, q16, q32, Percentage)


God_conception_wide <- pivot_wider(
  God_conception_selected,
  names_from = q32,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(God_conception_wide)[1] <- "God Conception"

# Create the table with kable and add styling with kableExtra
table <- kable(God_conception_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Conception of God among who believed in God" = 3, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Conception of God among who believed in God
God Conception	Both/Neither/Other (VOL.)	Don’t know/refused (VOL.)	God is a person	God is an impersonal force
Agnostic (not sure if there is a God)	5.909091	3.6363636	24.54545	65.909091
Buddhist	10.855263	3.9473684	23.35526	61.842105
Christian (VOL)	6.909091	5.4545455	64.72727	22.909091
Don’t know/Refused (VOL)	13.071895	15.0326797	31.37255	40.522876
Hindu	6.779661	2.5423729	30.93220	59.745763
Jehovah’s Witness (VOL)	3.365385	0.9615385	86.05769	9.615385
Jewish (Judaism)	4.821429	4.8214286	27.50000	62.857143
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	1.730104	0.8650519	92.38754	5.017301
Muslim (Islam)	8.333333	3.7037037	42.59259	45.370370
Nothing in particular	4.584626	5.4119269	41.84764	48.155808
Orthodox (Greek, Russian, or some other orthodox church)	6.936416	6.6473988	49.42197	36.994220
Protestant	4.188395	3.2841451	73.97648	18.550982
Roman Catholic	4.210258	4.0316407	64.53177	27.226333
Unitarian (Universalist) (VOL)	9.523810	6.3492063	23.01587	61.111111

Conclusion:

The data underscores the diversity of beliefs regarding the nature of God among individuals who identify with various religious traditions. While many perceive God as a personal entity, others conceptualize God as an impersonal force or hold alternative views. Understanding these diverse conceptions enriches our insight into the complex and multifaceted nature of religious beliefs across different faith communities.

Belief in God and Demographic Groups

Women demonstrate a significantly higher inclination than men to express absolute certainty in their belief in a personal God (81% vs. 72%).

God_belief_gender <- survey_data_2007 %>%
  filter(q30 == 'Yes') %>%
  group_by(sex) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(sex), !is.na(q31)) %>%  # Ensure q16 and q31 are not NA
  group_by(sex, q31) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(sex) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(sex, q31)


God_belief_gender_selected <- select(God_belief_gender, sex, q31, Percentage)


God_belief_gender_wide <- pivot_wider(
  God_belief_gender_selected,
  names_from = q31,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(God_belief_gender_wide)[1] <- "Sex"

# Create the table with kable and add styling with kableExtra
table <- kable(God_belief_gender_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Belief in God" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Belief in God
Sex	Absolutely certain	Don’t know/refused (VOL.)	Fairly certain	Not at all certain	Not too certain
Female	81.99158	0.7104990	14.72919	0.5465377	2.022189
Male	72.96798	0.7594417	21.35331	1.1904762	3.728790

Furthermore, older Americans are notably more likely than younger Americans to profess certain belief in a personal God. Among individuals aged 60 and above, almost eight-in-ten (81%) express this conviction, in contrast with (70%) of those under the age of 30.

God_belief_age <- filtered_data_age %>%
  filter(q30 == 'Yes') %>%
  group_by(sex) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(AgeGroup), !is.na(q31)) %>%  # Ensure q16 and q31 are not NA
  group_by(AgeGroup, q31) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(AgeGroup) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(AgeGroup, q31)


God_belief_age_selected <- select(God_belief_age, AgeGroup, q31, Percentage)


God_belief_age_wide <- pivot_wider(
  God_belief_age_selected,
  names_from = q31,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(God_belief_age_wide)[1] <- "AgeGroup"

# Create the table with kable and add styling with kableExtra
table <- kable(God_belief_age_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Belief in God" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Belief in God
AgeGroup	Absolutely certain	Don’t know/refused (VOL.)	Fairly certain	Not at all certain	Not too certain
10-20	69.57041	0.4773270	24.70167	1.1933174	4.057279
20-30	71.78884	0.5379960	22.86483	0.7733692	4.034970
30-40	75.30416	0.2774813	20.57631	0.8324440	3.009605
40-50	78.99427	0.4296626	17.23425	0.7638447	2.577976
50-60	77.52907	0.6395349	18.21221	0.8139535	2.805233
60-70	80.54747	0.7377208	15.41448	0.8542031	2.446127
70-80	82.28719	1.2611063	13.47091	0.9458298	2.034967
80+	79.96760	2.1058315	14.20086	0.8639309	2.861771

In general, Americans with a college education tend to be slightly less inclined to believe with certainty in a personal God compared to those without a degree.

God_belief_edu <- survey_data_2007 %>%
  group_by(educ) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(educ), !is.na(q21)) %>%  # Ensure q16 and q21 are not NA
  group_by(educ, q21) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(educ) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(educ, q21)


God_belief_edu_selected <- select(God_belief_edu, educ, q21, Percentage)


God_belief_edu_wide <- pivot_wider(
  God_belief_edu_selected,
  names_from = q21,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(God_belief_edu_wide)[1] <- "Education"

# Create the table with kable and add styling with kableExtra
table <- kable(God_belief_edu_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "How important is religion in your life" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	How important is religion in your life
Education	Don’t know/Refused (VOL.)	Not at all important	Not too important	Somewhat important	Very important
College graduate (B.S., B.A., or other 4-year degree)	0.7468658	8.015471	10.576154	26.79381	53.86770
Don’t know/Refused (VOL.)	8.9147287	4.263566	6.589147	23.25581	56.97674
High school graduate (Grade 12 or GED certificate)	0.7780922	5.254648	7.225142	26.35408	60.38804
High school incomplete (Grades 9-11)	1.2385321	4.724771	6.422018	22.98165	64.63303
None, or grade 1-8	1.7957351	3.030303	3.815937	16.94725	74.41077
Post-graduate training or professional schooling after colle	0.8466777	10.914780	11.522179	25.28990	51.42647
Some college, no 4-year degree (including associate degree)	0.9368244	6.665866	8.527504	26.00288	57.86692
Technical, trade, or vocational school AFTER high school	1.1173184	5.400372	7.355680	20.85661	65.27002

Afterlife

The data presents varying levels of certainty among individuals from different religious backgrounds regarding the existence of life after death:

• Absolutely Certain: A substantial proportion of respondents across several religious affiliations express absolute certainty about life after death. This belief is particularly strong among Mormons (90.86%), Jehovah’s Witnesses (88.89%), and Muslims (86.36%).

• Fairly Certain: Many respondents indicate a high degree of confidence, albeit not absolute certainty, in the existence of life after death. This sentiment is prevalent among Christians (Protestants: 76.48%, Christians: 73.73%, Catholics: 61.05%) and members of other religious groups.

• Uncertainty: Some respondents express varying degrees of uncertainty regarding life after death. This uncertainty is evident across different religious affiliations, with notable proportions of individuals falling into categories such as “Don’t know/refused” or “Not at all certain.”

Afterlife_belief <- survey_data_2007 %>%
  filter(q33 == 'Yes') %>% 
  group_by(q16) %>% 
  filter(n() > 50) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q34)) %>%  # Ensure q16 and q21 are not NA
  group_by(q16, q34) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q34)


Afterlife_belief_selected <- select(Afterlife_belief, q16, q34, Percentage)


Afterlife_belief_wide <- pivot_wider(
  Afterlife_belief_selected,
  names_from = q34,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Afterlife_belief_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Afterlife_belief_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Life After Death" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 


table

	Life After Death
Religion	Absolutely certain	Don’t know/refused (VOL.)	Fairly certain	Not at all certain	Not too certain
Agnostic (not sure if there is a God)	34.70149	0.7462687	37.686567	8.5820896	18.283582
Atheist (do not believe in God)	32.91139	2.5316456	43.037975	2.5316456	18.987342
Buddhist	47.77778	1.8518519	38.518518	2.9629630	8.888889
Christian (VOL)	73.73272	1.8433180	19.815668	1.3824885	3.225807
Don’t know/Refused (VOL)	52.52525	3.0303030	32.323232	4.0404040	8.080808
Hindu	50.60976	0.6097561	32.926829	4.2682927	11.585366
Jehovah’s Witness (VOL)	88.88889	1.1111111	6.666667	0.0000000	3.333333
Jewish (Judaism)	42.17391	0.8695652	37.826087	6.0869565	13.043478
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	90.86116	0.0000000	7.557118	0.0000000	1.581722
Muslim (Islam)	86.36364	1.1363636	9.090909	2.2727273	1.136364
Nothing in particular	46.72173	1.1874032	38.461539	2.3748064	11.254517
Orthodox (Greek, Russian, or some other orthodox church)	63.11787	0.7604563	31.178707	1.1406844	3.802281
Protestant	76.47719	0.6271869	19.654057	0.5017495	2.739816
Roman Catholic	61.05181	0.8634223	31.632653	0.8320251	5.620094
Unitarian (Universalist) (VOL)	41.42857	1.4285714	47.142857	1.4285714	8.571429

Conclusion:

The data highlights the diversity of beliefs concerning life after death among individuals from various religious traditions. While some express unwavering certainty, others hold more nuanced or uncertain perspectives. Understanding these diverse beliefs provides valuable insight into the complex and multifaceted nature of religious beliefs surrounding the concept of life after death.

Belief in Heaven and Hell

Belief in Heaven Across Different Religious Affiliations:

The data reflects the varying degrees of belief in heaven among individuals from diverse religious backgrounds:

• High Belief: A significant proportion of respondents across several religious groups express a strong belief in heaven. This belief is particularly prominent among Mormons (94.84%), Muslims (83.62%), and Protestants (81.96%).

• Moderate Belief: Some respondents exhibit a moderate level of belief in heaven. This category includes individuals from Orthodox Christian traditions (71.90%) and various Christian denominations.

• Low Belief: Certain religious affiliations demonstrate lower levels of belief in heaven compared to others. For instance, Buddhists (29.68%) and individuals identifying with “Nothing in particular” (45.40%) exhibit lower rates of belief in heaven.

• Uncertainty: A minority of respondents either express uncertainty about their belief in heaven or refrain from providing a clear response (“Don’t know/refused” category). This uncertainty is relatively low across most religious affiliations but is more pronounced among some groups like Jews (13.20%) and Atheists (3.50%).

Heaven_belief <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q35)) %>%  # Ensure q16 and q21 are not NA
  group_by(q16, q35) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q35)


Heaven_belief_selected <- select(Heaven_belief, q16, q35, Percentage)


Heaven_belief_wide <- pivot_wider(
  Heaven_belief_selected,
  names_from = q35,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Heaven_belief_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table_Heaven <- kable(Heaven_belief_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Belief in Heaven" = 3, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table_Heaven

	Belief in Heaven
Religion	Don’t know/refused (VOL.)	No	Other (VOL.)	Yes
Agnostic (not sure if there is a God)	11.743341	70.096852	2.663438	15.49637
Atheist (do not believe in God)	3.495146	85.242718	0.776699	10.48544
Buddhist	7.542579	59.854015	2.919708	29.68370
Christian (VOL)	8.561644	14.554794	5.308219	71.57534
Don’t know/Refused (VOL)	29.600000	38.800000	3.200000	28.40000
Hindu	10.894942	37.743191	3.112840	48.24903
Jehovah’s Witness (VOL)	3.755869	39.906103	9.389671	46.94836
Jewish (Judaism)	13.196481	52.346041	1.612903	32.84457
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	1.721170	2.409639	1.032702	94.83649
Muslim (Islam)	1.724138	12.068965	2.586207	83.62069
Nothing in particular	10.925277	41.111411	2.562719	45.40059
Orthodox (Greek, Russian, or some other orthodox church)	7.162534	17.630854	3.305785	71.90083
Protestant	4.860241	8.592601	4.587806	81.95935
Roman Catholic	5.654281	9.432087	1.851622	83.06201
Unitarian (Universalist) (VOL)	8.441558	72.727273	3.246753	15.58442

Hell_belief <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q36)) %>%  # Ensure q16 and q21 are not NA
  group_by(q16, q36) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q36)


Hell_belief_selected <- select(Hell_belief, q16, q36, Percentage)


Hell_belief_wide <- pivot_wider(
  Hell_belief_selected,
  names_from = q36,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Hell_belief_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table_Hell <- kable(Hell_belief_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Belief in Hell" = 3, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table_Hell

	Belief in Hell
Religion	Don’t know/refused (VOL.)	No	Other (VOL.)	Yes
Agnostic (not sure if there is a God)	9.685230	78.32930	2.1791768	9.806295
Atheist (do not believe in God)	2.912621	87.37864	0.9708738	8.737864
Buddhist	7.055961	70.31630	2.1897810	20.437956
Christian (VOL)	10.102740	27.91096	4.4520548	57.534247
Don’t know/Refused (VOL)	28.800000	49.20000	4.4000000	17.600000
Hindu	11.673152	50.19455	3.5019455	34.630350
Jehovah’s Witness (VOL)	1.408451	87.79343	1.8779343	8.920188
Jewish (Judaism)	9.237537	72.87390	0.8797654	17.008798
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	5.163511	29.43201	8.7779690	56.626506
Muslim (Islam)	3.448276	12.93103	5.1724138	78.448276
Nothing in particular	9.630429	55.48961	2.1850553	32.694902
Orthodox (Greek, Russian, or some other orthodox church)	12.672176	30.30303	3.5812672	53.443526
Protestant	7.769847	18.60186	3.2801177	70.348172
Roman Catholic	10.190133	27.22754	2.8209271	59.761402
Unitarian (Universalist) (VOL)	3.896104	88.96104	1.9480519	5.194805

Belief in Hell Across Different Religious Affiliations:

The dataset reveals a spectrum of beliefs regarding the concept of hell among individuals from various religious backgrounds:

• Strong Belief: A substantial proportion of respondents, particularly from Christian denominations like Jehovah’s Witnesses (8.92%) and Mormons (56.63%), express a strong belief in hell. This belief is also prevalent among Muslims (78.45%) and certain Protestant groups (70.35%).

• Moderate Belief: Some respondents demonstrate a moderate level of belief in hell, including individuals from Orthodox Christian traditions (53.44%) and Roman Catholics (59.76%).

• Low Belief: Certain religious affiliations exhibit lower levels of belief in hell compared to others. For instance, Buddhists (20.44%) and individuals identifying with “Nothing in particular” (32.69%) show lower rates of belief in hell.

• Uncertainty: A minority of respondents express uncertainty about their belief in hell or decline to provide a clear response (“Don’t know/refused” category). This uncertainty is relatively low across most religious affiliations but is more pronounced among some groups like Jews (9.24%) and Atheists (2.91%).

Combined_Heaven_Hell_wide = bind_cols (Heaven_belief_wide$Religion , Heaven_belief_wide$Yes , Hell_belief_wide$Yes)

## New names:
## • `` -> `...1`
## • `` -> `...2`
## • `` -> `...3`

colnames(Combined_Heaven_Hell_wide)[1] <- "Religion"
colnames(Combined_Heaven_Hell_wide)[2] <- "Heaven"
colnames(Combined_Heaven_Hell_wide)[3] <- "Hell"


# Reshape the data to a long format
long_data <- pivot_longer(
  Combined_Heaven_Hell_wide,
  cols = c("Heaven", "Hell"),
  names_to = "Belief",
  values_to = "Percentage"
)

# Create the bar plot
p <- ggplot(long_data, aes(x = Religion, y = Percentage, fill = Belief)) +
  geom_bar(stat = "identity", position = position_dodge(width = 0.7)) +
  labs(title = "Percentage of People Believing in Heaven and Hell by Religion",
       x = "Religion",
       y = "Percentage",
       fill = "Belief") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1),
        legend.position = "bottom")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 700)

# Show the interactive plot in your default browser

interactive_plot

Miracles and the Supernatural

Belief in the Occurrence of Miracles Today:

The survey data provides insights into the beliefs regarding the occurrence of miracles in contemporary times compared to ancient eras among individuals from various religious backgrounds:

• Strong Belief: A significant proportion of respondents from the Mormon Church (LDS) (79.17%), Protestant denominations (53.26%), and certain Christian denominations (52.05% to 46.80%) express strong agreement that miracles still occur today, akin to ancient times. This belief is notably high among Mormons, indicating a firm conviction in the continued manifestation of miracles.

• Moderate Belief: Some respondents, particularly from Orthodox Christian traditions (41.87%), Islam (39.66%), and Hinduism (24.12%), express a moderate level of agreement with the notion that miracles still happen today. While not as pronounced as in the aforementioned groups, this belief remains notable within these religious affiliations.

• Uncertainty: A portion of respondents either express uncertainty about the occurrence of miracles in contemporary times or decline to provide a definitive stance (“Don’t know/refused” category). This uncertainty varies across religious affiliations but is particularly notable among Atheists (61.75%) and individuals identifying with “Nothing in particular” (17.80%).

• Disagreement: Certain respondents, including Atheists, Agnostics, and individuals from Jewish, Unitarian Universalist, and certain Christian denominations, express either complete disagreement or a majority disagreement with the idea that miracles still occur today as they did in ancient times.

Miracle_belief <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q39a)) %>%  # Ensure q16 and q39b are not NA
  group_by(q16, q39a) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q39a)


Miracle_belief_selected <- select(Miracle_belief, q16, q39a, Percentage)


Miracle_belief_wide <- pivot_wider(
  Miracle_belief_selected,
  names_from = q39a,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Miracle_belief_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table_Miracle <- kable(Miracle_belief_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Question: Miracles still occur today as in ancient times" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table_Miracle

	Question: Miracles still occur today as in ancient times
Religion	Completely agree	Completely disagree	Don’t know/refused (VOL.)	Mostly agree	Mostly disagree
Agnostic (not sure if there is a God)	13.075060	36.3196126	4.6004843	21.428571	24.576271
Atheist (do not believe in God)	9.514563	61.7475728	3.8834951	9.126214	15.728155
Buddhist	30.170316	16.7883212	5.3527981	28.953771	18.734793
Christian (VOL)	52.054794	5.6506849	5.9931507	28.938356	7.363014
Don’t know/Refused (VOL)	22.400000	16.8000000	22.4000000	24.000000	14.400000
Hindu	24.124514	12.4513619	4.6692607	43.579766	15.175097
Jehovah’s Witness (VOL)	14.084507	46.4788732	5.1643192	15.492958	18.779343
Jewish (Judaism)	18.768328	24.6334311	6.1583578	29.618768	20.821114
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	79.173838	0.8605852	0.3442341	17.383821	2.237522
Muslim (Islam)	39.655172	12.9310345	6.0344828	27.586207	13.793103
Nothing in particular	27.569463	17.8041543	4.5049906	32.560022	17.561370
Orthodox (Greek, Russian, or some other orthodox church)	41.873278	8.5399449	2.7548209	36.639119	10.192837
Protestant	53.261047	3.8304364	2.7352476	32.163679	8.009590
Roman Catholic	46.800050	4.2127501	2.9824779	36.858457	9.146266
Unitarian (Universalist) (VOL)	14.935065	23.3766234	5.1948052	36.363636	20.129870

Similarly, about two-thirds of U.S. adults (68%) believe in the activity of angels and demons in the world. This belief is especially common among members of Christian traditions, including approximately nine-in-ten members of evangelical Protestant churches, Jehovah’s Witnesses, and Mormons.

DEMON_belief <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q39b)) %>%  # Ensure q16 and q21 are not NA
  group_by(q16, q39b) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q39b)


DEMON_belief_selected <- select(DEMON_belief, q16, q39b, Percentage)


DEMON_belief_wide <- pivot_wider(
  DEMON_belief_selected,
  names_from = q39b,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(DEMON_belief_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table_Demon <- kable(DEMON_belief_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Question: Angels and demons are active in the world" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table_Demon

	Question: Angels and demons are active in the world
Religion	Completely agree	Completely disagree	Don’t know/refused (VOL.)	Mostly agree	Mostly disagree
Agnostic (not sure if there is a God)	6.416465	55.690073	3.3898305	11.016949	23.4866828
Atheist (do not believe in God)	5.242718	76.893204	3.3009709	6.407767	8.1553398
Buddhist	12.165450	35.766423	6.3260341	22.627737	23.1143552
Christian (VOL)	44.863014	8.219178	9.7602740	27.568493	9.5890411
Don’t know/Refused (VOL)	16.800000	28.000000	25.2000000	13.600000	16.4000000
Hindu	15.175097	35.019455	7.7821012	19.455253	22.5680934
Jehovah’s Witness (VOL)	78.873239	2.816901	0.9389671	16.901408	0.4694836
Jewish (Judaism)	7.478006	54.692082	7.0381232	10.703812	20.0879765
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	58.864027	3.270224	3.7865749	29.948365	4.1308090
Muslim (Islam)	45.689655	6.034483	8.6206897	31.034483	8.6206897
Nothing in particular	19.800378	29.646614	6.4472619	24.089560	20.0161856
Orthodox (Greek, Russian, or some other orthodox church)	32.231405	13.774105	5.5096419	32.231405	16.2534435
Protestant	48.329973	7.350297	4.6749850	29.014330	10.6304146
Roman Catholic	33.962968	10.289549	5.7909780	34.708587	15.2479185
Unitarian (Universalist) (VOL)	5.844156	52.597403	5.8441558	10.389610	25.3246753

In contrast, belief in the activity of angels and demons is less common among Buddhists and Hindus, as well as among Jews. While relatively few atheists and agnostics hold this belief.

Religious Practices

Worship Attendance

The data illustrates diverse patterns of attendance across different religious groups. For instance, some groups show higher rates of frequent attendance, such as Jehovah’s Witnesses and Mormons, with a significant proportion attending religious services several times a day. Within the Christian category, there’s notable variation in attendance habits. For example, while a considerable portion of Roman Catholics attend services a few times a week or more, Protestant attendees tend to spread more evenly across different levels of attendance frequency.

Worship_frequency <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q41)) %>%  # Ensure q16 and q41 are not NA
  group_by(q16, q41) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q41)


Worship_frequency_selected <- select(Worship_frequency, q16, q41, Percentage)


Worship_frequency_wide <- pivot_wider(
  Worship_frequency_selected,
  names_from = q41,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Worship_frequency_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Worship_frequency_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Attendance at Religious Services" = 7, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Attendance at Religious Services
Religion	A few times a month	A few times a week	Don’t know/Refused (VOL.)	Never	Once a day	Once a week	Seldom	Several times a day
Agnostic (not sure if there is a God)	6.537530	6.295400	0.3631961	44.4309927	3.631961	1.8159806	30.871671	6.053269
Atheist (do not believe in God)	1.553398	2.330097	2.1359223	79.2233010	2.330097	1.5533981	8.737864	2.135922
Buddhist	8.272506	10.948905	1.4598540	15.3284672	19.221411	1.9464720	15.328467	27.493917
Christian (VOL)	3.767123	14.212329	6.6780822	2.5684932	18.664384	1.3698630	7.191781	45.547945
Don’t know/Refused (VOL)	4.400000	7.600000	20.0000000	19.2000000	9.200000	1.2000000	20.800000	17.600000
Hindu	5.058366	10.894942	1.9455253	4.6692607	36.186770	3.1128405	12.840467	25.291829
Jehovah’s Witness (VOL)	1.877934	5.164319	0.4694836	0.4694836	9.859155	0.4694836	1.408451	80.281690
Jewish (Judaism)	8.357771	13.196481	3.5190616	18.0351906	12.023460	4.5454545	27.712610	12.609971
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	2.753873	6.884682	0.6884682	0.3442341	16.006885	1.2048193	4.130809	67.986231
Muslim (Islam)	0.862069	6.034483	0.8620690	7.7586207	5.172414	2.5862069	8.620690	68.103448
Nothing in particular	7.661181	11.788508	2.1041273	24.2514162	11.087132	2.1580793	24.925816	16.023739
Orthodox (Greek, Russian, or some other orthodox church)	5.234160	15.151515	1.9283747	4.1322314	26.997245	3.0303030	11.570248	31.955923
Protestant	4.380755	14.166621	1.3839699	1.4602517	21.500572	2.1631341	6.195172	48.749523
Roman Catholic	6.288058	16.751584	1.1805642	2.3984093	25.599602	3.6908165	9.282963	34.808003
Unitarian (Universalist) (VOL)	4.545454	11.038961	3.2467532	22.0779221	14.935065	5.1948052	18.831169	20.129870

Atheists and agnostics exhibit lower rates of religious service attendance compared to religious groups. The majority of respondents in these categories report never attending religious services, with atheists (79.2%) and agnostics (44.4%) being the highest in this regard.Those who identify as having “nothing in particular” also show varied attendance patterns, with a significant proportion reporting infrequent or no attendance. Specifically, 24.3% report never attending religious services..Minority religious groups such as Buddhists, Hindus, Muslims, and Jews display diverse attendance patterns, with some members attending services frequently while others rarely do. For example, among Muslims, 68.1% report attending services several times a day, while among Jews, 27.7% attend services a few times a week or more.

Religious Upbringing of Children

Overall, nearly two-thirds of adults (63%) who have children under age 18 living at home say they pray or read Scripture with their children. Mormons (91%) are especially likely to do this. Roughly eight-in-ten members of protestan (81%) and historically black (77%) churches also pray or read Scripture with their children. Jews (41%), Buddhists (37%) and the unaffiliated (31%) are much less likely to pray or read Scripture with their children compared with members of many other religious traditions. But there are major differences within the unaffiliated group; a majority of the religious unaffiliated (52%) pray or read Scripture with their children, compared with only 16% of the secular unaffiliated, 19% of agnostics and 8% of atheists. Similar patterns are seen in parents’ decisions about sending their children to Sunday school or other religious education programs. Six-in-ten Americans with children under age 18 living at home arrange for them to attend such programs. Mormons (90%) stand out for their propensityto enroll their children in religious education, followed by members of evangelical (79%) and historically black (77%) churches. More than a third of the unaffiliated (35%) send their children to religious education programs, including a quarter of atheists (24%) and about half of the religious unaffiliated (49%). Among parents with children under age 18 living in their homes, 15% home-school their children or send them to a religious school instead of a public school. Jews (27%) and Orthodox Christians (30%) are most likely to do this.

Upbringing_children <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q55a)) %>%  # Ensure q16 and q41 are not NA
  group_by(q16, q55a) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q55a)


Upbringing_children_selected <- select(Upbringing_children, q16, q55a, Percentage)


Upbringing_children_wide <- pivot_wider(
  Upbringing_children_selected,
  names_from = q55a,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Upbringing_children_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table_1 <- kable(Upbringing_children_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Send any of your children to Sunday school" = 2, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table_1

	Send any of your children to Sunday school
Religion	No	Yes	Don’t know/Refused (VOL.)
Agnostic (not sure if there is a God)	81.025641	18.974359	0.0000000
Atheist (do not believe in God)	90.163934	7.377049	2.4590164
Buddhist	65.833333	34.166667	0.0000000
Christian (VOL)	28.947368	67.894737	3.1578947
Don’t know/Refused (VOL)	54.929578	36.619718	8.4507042
Hindu	34.375000	64.062500	1.5625000
Jehovah’s Witness (VOL)	13.888889	86.111111	0.0000000
Jewish (Judaism)	59.880240	40.119760	0.0000000
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	8.433735	91.566265	0.0000000
Muslim (Islam)	44.642857	55.357143	0.0000000
Nothing in particular	65.822785	33.502110	0.6751055
Orthodox (Greek, Russian, or some other orthodox church)	45.283019	54.716981	0.0000000
Protestant	25.134823	74.345146	0.5200308
Roman Catholic	36.720209	62.831528	0.4482630
Unitarian (Universalist) (VOL)	61.363636	38.636364	0.0000000

Upbringing_children_2 <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q55b)) %>%  # Ensure q16 and q41 are not NA
  group_by(q16, q55b) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q55b)


Upbringing_children_2_selected <- select(Upbringing_children_2, q16, q55b, Percentage)


Upbringing_children_2_wide <- pivot_wider(
  Upbringing_children_2_selected,
  names_from = q55b,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Upbringing_children_2_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table_2 <- kable(Upbringing_children_2_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Pray or read the Scripture with your child" = 2, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table_2

	Pray or read the Scripture with your child
Religion	No	Yes	Don’t know/Refused (VOL.)
Agnostic (not sure if there is a God)	68.205128	31.79487	0.0000000
Atheist (do not believe in God)	72.950820	23.77049	3.2786885
Buddhist	72.500000	27.50000	0.0000000
Christian (VOL)	40.526316	57.36842	2.1052632
Don’t know/Refused (VOL)	57.746479	35.21127	7.0422535
Hindu	75.781250	21.87500	2.3437500
Jehovah’s Witness (VOL)	63.888889	34.72222	1.3888889
Jewish (Judaism)	43.113773	56.88623	0.0000000
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	8.835341	91.16466	0.0000000
Muslim (Islam)	58.928571	41.07143	0.0000000
Nothing in particular	62.278481	37.21519	0.5063291
Orthodox (Greek, Russian, or some other orthodox church)	39.622642	60.37736	0.0000000
Protestant	24.229584	75.28891	0.4815100
Roman Catholic	43.332088	56.10758	0.5603287
Unitarian (Universalist) (VOL)	36.363636	63.63636	0.0000000

Upbringing_children_3 <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q55c)) %>%  # Ensure q16 and q41 are not NA
  group_by(q16, q55c) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q55c)


Upbringing_children_3_selected <- select(Upbringing_children_3, q16, q55c, Percentage)


Upbringing_children_3_wide <- pivot_wider(
  Upbringing_children_3_selected,
  names_from = q55c,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Upbringing_children_3_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table_3 <- kable(Upbringing_children_3_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Send your child/childern to religious school instead of a public school" = 2, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table_3

	Send your child/childern to religious school instead of a public school
Religion	Don’t know/Refused (VOL.)	No	Yes
Agnostic (not sure if there is a God)	0.5128205	90.76923	8.717949
Atheist (do not believe in God)	2.4590164	91.80328	5.737705
Buddhist	0.8333333	89.16667	10.000000
Christian (VOL)	2.1052632	81.57895	16.315790
Don’t know/Refused (VOL)	7.0422535	80.28169	12.676056
Hindu	1.5625000	93.75000	4.687500
Jehovah’s Witness (VOL)	1.3888889	84.72222	13.888889
Jewish (Judaism)	0.0000000	76.64671	23.353293
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	0.0000000	93.97590	6.024096
Muslim (Islam)	0.0000000	85.71429	14.285714
Nothing in particular	0.5907173	92.32068	7.088608
Orthodox (Greek, Russian, or some other orthodox church)	0.0000000	68.86792	31.132075
Protestant	0.8859784	83.82126	15.292758
Roman Catholic	0.7097497	77.54950	21.740755
Unitarian (Universalist) (VOL)	0.0000000	90.90909	9.090909

Combined_Upbringing_children_wide = bind_cols (Upbringing_children_wide$Religion , Upbringing_children_wide$Yes , Upbringing_children_2_wide$Yes , Upbringing_children_3_wide$Yes)

## New names:
## • `` -> `...1`
## • `` -> `...2`
## • `` -> `...3`
## • `` -> `...4`

colnames(Combined_Upbringing_children_wide)[1] <- "Religion"
colnames(Combined_Upbringing_children_wide)[2] <- "Pray and read scripture"
colnames(Combined_Upbringing_children_wide)[3] <- "attend religious school on sunday"
colnames(Combined_Upbringing_children_wide)[4] <- "attend religious school instead of public school"


# Reshape the data to a long format
long_data <- pivot_longer(
  Combined_Upbringing_children_wide,
  cols = c("Pray and read scripture", "attend religious school on sunday","attend religious school instead of public school"),
  names_to = "Belief",
  values_to = "Percentage"
)

# Create the bar plot
p <- ggplot(long_data, aes(x = Religion, y = Percentage, fill = Belief)) +
  geom_bar(stat = "identity", position = position_dodge(width = 0.7)) +
  labs(title = "Percentage of People saying yes to the quesions",
       x = "Religion",
       y = "Percentage",
       fill = "Belief") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1),
        legend.position = "bottom")


# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 900)

# Show the interactive plot in your default browser

interactive_plot

Similar patterns are seen in parents’ decisions about sending their children to Sunday school or other religious education programs. Six-in-ten Americans with children under age 18 living at home arrange for them to attend such programs. Mormons (90%) stand out for their propensity to enroll their children in religious education. More than a third of the unaffiliated (35%) send their children to religious education programs, including a quarter of atheists (24%). Among parents with children under age 18 living in their homes, 15% home-school their children or send them to a religious school instead of a public school. Jews (27%) and Orthodox Christians (30%) are most likely to do this.

Spiritual Experiences

Spritual Peace and Wonder

A majority of Americans (52%) report experiencing a sense of spiritual peace and well-being at least once a week. High rates of weekly spiritual peace are observed among Jehovah’s Witnesses and Mormons (77% and 71%, respectively), Muslims (64%). Conversely, Jews (38%) and the unaffiliated (35%) are among the groups least likely to experience spiritual peace on a weekly basis.

In contrast, a smaller proportion of Americans (39%) report feeling a deep sense of wonder about the universe on a weekly basis. Notably, over half of Buddhists (57%) and Muslims (53%) report experiencing this deep sense of wonder weekly, along with nearly half of Jehovah’s Witnesses (49%).

Peace_Wonder <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q43a)) %>%  # Ensure q16 and q43a are not NA
  group_by(q16, q43a) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q43a)


Peace_Wonder_selected <- select(Peace_Wonder, q16, q43a, Percentage)


Peace_Wonder_wide <- pivot_wider(
  Peace_Wonder_selected,
  names_from = q43a,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Peace_Wonder_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Peace_Wonder_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "feel a deep sense of spiritual peace and well-being" = 5, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	feel a deep sense of spiritual peace and well-being
Religion	At least once a week	Don’t know/Refused (VOL.)	Never	Once or twice a month	Seldom	Several times a year
Agnostic (not sure if there is a God)	32.32446	3.510896	14.5278450	14.769976	20.217918	14.648910
Atheist (do not believe in God)	29.51456	6.213592	30.2912621	9.320388	15.145631	9.514563
Buddhist	57.66423	3.649635	4.6228710	14.841849	9.975669	9.245742
Christian (VOL)	56.67808	7.191781	3.9383562	11.301370	10.445206	10.445206
Don’t know/Refused (VOL)	40.40000	19.200000	11.6000000	7.200000	13.600000	8.000000
Hindu	46.69261	4.669261	8.1712062	7.392996	10.505837	22.568093
Jehovah’s Witness (VOL)	79.34272	2.816901	3.2863850	5.633803	2.816901	6.103286
Jewish (Judaism)	35.33724	6.011730	9.6774194	14.076246	21.114369	13.782991
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	74.01033	2.237522	0.5163511	11.187608	4.475043	7.573150
Muslim (Islam)	65.51724	3.448276	3.4482759	12.931035	7.758621	6.896552
Nothing in particular	37.25384	4.343135	13.2452118	12.274076	20.852441	12.031292
Orthodox (Greek, Russian, or some other orthodox church)	46.83196	3.581267	5.7851240	16.253443	12.121212	15.426997
Protestant	61.29788	3.220182	2.8932600	13.262137	8.483627	10.842914
Roman Catholic	48.72623	3.442277	4.7719647	15.558593	12.016901	15.484031
Unitarian (Universalist) (VOL)	54.54545	1.948052	5.8441558	18.181818	11.038961	8.441558

Peace_Wonder <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q43a)) %>%  # Ensure q16 and q43b are not NA
  group_by(q16, q43b) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q43b)


Peace_Wonder_selected <- select(Peace_Wonder, q16, q43b, Percentage)


Peace_Wonder_wide <- pivot_wider(
  Peace_Wonder_selected,
  names_from = q43b,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Peace_Wonder_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Peace_Wonder_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "feel a deep sense of wonder about the universe" = 5, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	feel a deep sense of wonder about the universe
Religion	At least once a week	Don’t know/Refused (VOL.)	Never	Once or twice a month	Seldom	Several times a year
Agnostic (not sure if there is a God)	47.82082	0.4842615	6.537530	16.828087	15.254237	13.075060
Atheist (do not believe in God)	39.80583	2.7184466	15.728155	13.203883	16.310680	12.233010
Buddhist	61.55718	2.9197080	7.299270	11.678832	7.542579	9.002433
Christian (VOL)	41.09589	6.8493151	12.500000	12.157534	14.897260	12.500000
Don’t know/Refused (VOL)	44.00000	16.8000000	8.800000	9.200000	14.400000	6.800000
Hindu	37.35409	3.5019455	7.782101	12.062257	14.396887	24.902724
Jehovah’s Witness (VOL)	52.58216	5.6338028	16.901408	9.389671	7.042254	8.450704
Jewish (Judaism)	39.14956	3.9589443	11.583578	12.903226	15.249267	17.155425
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	43.71773	3.4423408	8.950086	16.006885	13.941480	13.941480
Muslim (Islam)	54.31034	5.1724138	6.896552	12.931035	13.793103	6.896552
Nothing in particular	39.06124	3.0213110	11.141084	12.651740	19.773402	14.351227
Orthodox (Greek, Russian, or some other orthodox church)	36.91460	3.8567493	11.019284	13.223140	15.151515	19.834711
Protestant	41.02327	4.5769084	10.292595	14.139378	16.460524	13.507328
Roman Catholic	34.91985	3.8275134	10.960606	13.992792	19.572512	16.726730
Unitarian (Universalist) (VOL)	66.88312	1.9480519	2.597403	12.337662	7.792208	8.441558

While Mormons and protestant diverge from the general public on various religious beliefs and practices, they align closely with the broader population on this particular question. Approximately 41% of protestant, 43% of Mormons report feeling a deep sense of wonder about the universe on a weekly basis. This places them in close proximity to Jews (40%) and the unaffiliated (39%).

Interestingly, within the unaffiliated category, agnostics stand out as the subgroup most likely to frequently experience a sense of wonder, with 48% reporting weekly occurrences. This contrasts with the figures for atheists (37%), the religious unaffiliated (40%).

Is There More Than One Path to Salvation?

Quesion: My religion is the one true faith leading to eternal life, OR Many religions can lead to eternal life.

Faith_flexibility <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q40a)) %>%  # Ensure q16 and q43b are not NA
  group_by(q16, q40a) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q40a)


Faith_flexibility_selected <- select(Faith_flexibility, q16, q40a, Percentage)


Faith_flexibility_wide <- pivot_wider(
  Faith_flexibility_selected,
  names_from = q40a,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Faith_flexibility_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Faith_flexibility_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "is There More Than One Path to Salvation" = 3, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	is There More Than One Path to Salvation
Religion	Don’t know/refused (VOL)	Many religions can lead to eternal life - Second statement	My religion is the one, true faith leading to eternal life -	Neither/Both equally (VOL)
Buddhist	3.892944	85.88808	3.8929440	6.3260341
Christian (VOL)	8.047945	66.26712	23.4589041	2.2260274
Hindu	4.669261	89.49416	4.2801556	1.5564202
Jehovah’s Witness (VOL)	2.816901	14.55399	82.1596244	0.4694836
Jewish (Judaism)	6.158358	82.40469	3.5190616	7.9178886
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	1.549053	39.07057	56.6265060	2.7538726
Muslim (Islam)	10.344828	56.89655	29.3103448	3.4482759
Orthodox (Greek, Russian, or some other orthodox church)	4.958678	71.62534	20.1101928	3.3057851
Protestant	4.021141	67.88536	25.6906228	2.4028769
Roman Catholic	2.796073	81.91873	13.4211507	1.8640487
Unitarian (Universalist) (VOL)	3.246753	81.81818	0.6493506	14.2857143

# Assuming Combined_Heaven_Hell_wide has two columns: Religion, and Percentage with values for Heaven and Hell
# Inverting sign for "Hell" percentages
# Sample data frame
religions <- Faith_flexibility_wide$Religion
many_religions <- Faith_flexibility_wide$`Many religions can lead to eternal life - Second statement`
one_true_faith <- Faith_flexibility_wide$`My religion is the one, true faith leading to eternal life -`

df <- data.frame(
  Religion = factor(religions, levels = religions),
  `who_say_many_religions_can_lead_to_eternal_life` = many_religions,
  `who_say_my_religion_is_the_one_true_faith_leading_to_eternal_life` = -one_true_faith
)


# Sort the religions based on 'Many religions can lead to eternal life'
df_sorted <- df %>%
  arrange(desc(`who_say_many_religions_can_lead_to_eternal_life`))

# Update the factor levels of Religion to the new order
df_sorted$Religion <- factor(df_sorted$Religion, levels = df_sorted$Religion)

long_df <- pivot_longer(
  df_sorted,
  cols = c(`who_say_many_religions_can_lead_to_eternal_life`, `who_say_my_religion_is_the_one_true_faith_leading_to_eternal_life`),
  names_to = "Belief",
  values_to = "Percentage"
)

long_df <- long_df[order(long_df$Percentage, decreasing = TRUE),]

# Create the plot
p <- ggplot(long_df, aes(x = Religion, y = Percentage, fill = Belief)) +
  geom_bar(stat = "identity", position = "identity") +
  coord_flip() +
  scale_fill_manual(values = c("#FFA07A", "blue")) +
  labs(title = "Religious Exclusivity",
       x = "",
       y = "Percentage")+
  theme_minimal() +
  theme(legend.position = "right",
        
        axis.ticks.y = element_blank(),
        panel.grid = element_blank())


# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 500)

# Show the interactive plot in your default browser

interactive_plot

The majority of Americans with a religious affiliation believe that many religions can lead to eternal life, with seven-in-ten holding this view. This perspective is widespread across various religious traditions, including more than eight-in-ten Jews, Buddhists, Hindus, and Protestant, as well as nearly eight-in-ten Catholics. However 56% of Muslims, share this belief. Notably, among Mormons and Jehovah’s Witnesses, a majority assert that their religion is the sole path to eternal life.

How Strictly to Interpret the Faith?

Quesion: There is only ONE true way to interpret the teachings of my religion, OR There is MORE than one true way to interpret the teachings of my religion.

More than two-thirds of Americans with a religious affiliation (68%) believe in the possibility of multiple interpretations of their faith’s teachings. A majority of members of evangelical churches (64%) share this view, along with upwards of six-in-ten individuals from most other religious traditions. Notably, a significant proportion of Jews (89%) and Buddhists (90%) express the belief in multiple valid interpretations of their religion’s teachings. However, Mormons and Jehovah’s Witnesses differ on this point, with majorities from both groups (54% among Mormons, 77% among Jehovah’s Witnesses) asserting that there is only one true interpretation of their religion’s teachings.

Faith_flexibility <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q40b)) %>%  # Ensure q16 and q40b are not NA
  group_by(q16, q40b) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q40b)


Faith_flexibility_selected <- select(Faith_flexibility, q16, q40b, Percentage)


Faith_flexibility_wide <- pivot_wider(
  Faith_flexibility_selected,
  names_from = q40b,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Faith_flexibility_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Faith_flexibility_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Strictness of Interpretation" = 3, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Strictness of Interpretation
Religion	Don’t know/refused (VOL)	Neither/Both equally (VOL)	There is MORE than one true way to interpret the teachings o	There is only ONE true way to interpret the teachings of my
Buddhist	3.406326	0.4866180	90.99757	5.109489
Christian (VOL)	7.363014	2.7397260	64.72603	25.171233
Hindu	3.891051	1.1673152	84.82490	10.116732
Jehovah’s Witness (VOL)	4.225352	1.4084507	14.55399	79.812207
Jewish (Judaism)	3.519062	0.7331378	90.90909	4.838710
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	2.409639	1.5490534	39.75904	56.282272
Muslim (Islam)	6.896552	1.7241379	62.93103	28.448276
Orthodox (Greek, Russian, or some other orthodox church)	4.407713	1.1019284	67.49311	26.997245
Protestant	3.863129	1.1115349	64.05492	30.970414
Roman Catholic	3.380142	0.7704735	77.27103	18.578352
Unitarian (Universalist) (VOL)	1.298701	2.5974026	92.85714	3.246753

religions <- Faith_flexibility_wide$Religion
many_religions <- Faith_flexibility_wide$`There is MORE than one true way to interpret the teachings o`
one_true_faith <- Faith_flexibility_wide$`There is only ONE true way to interpret the teachings of my`

df <- data.frame(
  Religion = factor(religions, levels = religions),
  `There_is_MORE_than_one_true_way_to_interpret_the_teachings` = many_religions,
  `There_is_only_one_true_way_to_interpret_the_teachings` = -one_true_faith
)


# Sort the religions based on 'Many religions can lead to eternal life'
df_sorted <- df %>%
  arrange(desc(`There_is_MORE_than_one_true_way_to_interpret_the_teachings`))

# Update the factor levels of Religion to the new order
df_sorted$Religion <- factor(df_sorted$Religion, levels = df_sorted$Religion)

long_df <- pivot_longer(
  df_sorted,
  cols = c(`There_is_MORE_than_one_true_way_to_interpret_the_teachings`, `There_is_only_one_true_way_to_interpret_the_teachings`),
  names_to = "Belief",
  values_to = "Percentage"
)

long_df <- long_df[order(long_df$Percentage, decreasing = TRUE),]

# Create the plot
p <- ggplot(long_df, aes(x = Religion, y = Percentage, fill = Belief)) +
  geom_bar(stat = "identity", position = "identity") +
  coord_flip() +
  scale_fill_manual(values = c("#FFA07A", "blue")) +
  labs(title = "Religious Interpretablility",
       x = "",
       y = "Percentage")+
  theme_minimal() +
  theme(legend.position = "right",
        
        axis.ticks.y = element_blank(),
        panel.grid = element_blank())


# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 500)

# Show the interactive plot in your default browser

interactive_plot

Beliefs About Morality

Quesion: There are clear and absolute standards for what is right and wrong.

Right_Wrong <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q10b)) %>%  # Ensure q16 and q10b are not NA
  group_by(q16, q10b) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q10b)


Right_Wrong_selected <- select(Right_Wrong, q16, q10b, Percentage)


Right_Wrong_wide <- pivot_wider(
  Right_Wrong_selected,
  names_from = q10b,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Right_Wrong_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Right_Wrong_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Absolute Standards of Right and Wrong" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Absolute Standards of Right and Wrong
Religion	Completely agree	Completely disagree	Don’t know/Refused (VOL.)	Mostly agree	Mostly disagree
Agnostic (not sure if there is a God)	17.79661	12.469734	2.058111	40.43584	27.239709
Atheist (do not believe in God)	19.02913	14.563107	2.330097	38.05825	26.019417
Buddhist	13.38200	20.924574	3.163017	35.27981	27.250608
Christian (VOL)	38.69863	7.020548	3.938356	37.15753	13.184931
Don’t know/Refused (VOL)	26.80000	13.600000	8.800000	29.60000	21.200000
Hindu	20.62257	9.338521	6.225681	43.19066	20.622568
Jehovah’s Witness (VOL)	58.21596	4.225352	2.816901	26.76056	7.981221
Jewish (Judaism)	18.03519	9.824047	2.492669	44.28152	25.366569
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	45.78313	3.098107	1.549053	41.65232	7.917384
Muslim (Islam)	33.62069	7.758621	5.172414	36.20690	17.241379
Nothing in particular	29.24197	9.873213	2.346911	38.89938	19.638522
Orthodox (Greek, Russian, or some other orthodox church)	31.95592	7.988981	4.683196	40.22039	15.151515
Protestant	42.73961	4.811203	2.157685	38.13545	12.156051
Roman Catholic	36.47322	5.778551	2.423263	42.15235	13.172611
Unitarian (Universalist) (VOL)	11.03896	14.285714	4.545454	37.01299	33.116883

While most Americans believe in absolute standards of right and wrong, there are some differences among groups. For example, Buddhists, atheists, and agnostics are somewhat less likely to believe in absolute standards compared to other religious groups or the religiously unaffiliated as a whole. This suggests that while there’s a general trend toward believing in universal moral principles, individual beliefs can vary based on religious background or lack thereof.

Conclusion

The analysis of religious beliefs and practices across a diverse array of faiths and those without religious affiliation reveals significant insights into how religion—or the absence of it—shapes individuals’ worldviews, ethical standards, and daily practices.

Religious Affiliations: For those with religious affiliations, religion typically serves as a foundational element in their lives, deeply influencing their morals, behaviors, and outlooks. These individuals often view their religious tenets as not just beliefs but as integral truths that guide their understanding of the world, their behavior, and their expectations of the afterlife. They tend to participate more actively in community and religious rituals, which reinforces their faith and fosters a sense of belonging and identity. The belief in divine beings, miracles, and an active spiritual realm is prevalent, reflecting a worldview where the divine intimately interacts with the world.

Non-religious Individuals (Atheists, Agnostics, Unaffiliated): In contrast, those without religious affiliations often embrace a worldview grounded in secularism or humanism. This group tends to place less importance on structured beliefs and shows a higher propensity for skepticism towards supernatural claims and absolute truths. However, this does not imply a lack of ethical or moral consideration; many non-religious individuals still exhibit strong moral convictions and a sense of wonder about the universe, which they might attribute to philosophical or humanistic insights rather than spiritual or religious beliefs. Importantly, their engagement with spiritual practices, while less frequent, often occurs in a context that values cultural heritage or personal exploration over religious duty.

Shared Human Experience: Despite these differences, both groups share fundamental human experiences and emotions, such as the quest for meaning, the experience of awe, and the need for ethical guidelines. Both religious and non-religious individuals seek peace, purpose, and explanations for the complex world around them, albeit through different lenses.

Conclusion: This analysis suggests that while religious beliefs significantly shape one’s worldview and daily practices, underlying human experiences such as ethical dilemmas, spiritual peace, and existential wonder are universal. These findings indicate that the essence of human experience crosses the boundaries of religious and secular divides, pointing to a shared search for understanding and meaning that transcends specific belief systems. Thus, the dialogue between individuals of varying religious and non-religious backgrounds is not only possible but essential for a comprehensive understanding of the human condition.

Chapter 2: Social and Political Views

While religion might not always be seen as the main influence on social and political views for most Americans, the Landscape Survey reveals strong connections between religious affiliation, beliefs, and societal attitudes. This suggests that religion might have a significant, albeit indirect, role in shaping people’s perspectives more than they realize.

This chapter delves into the social and political stances of different religious groups. When possible, it breaks down these traditions further based on key measures of religious commitment, like how often someone attends worship services or the importance they place on religion in their lives. The analysis reveals that religious divisions often align with political fault lines, but also cut across them.

Despite the diversity of religious beliefs and practices in the U.S., there’s widespread agreement across most faiths about the positive impact of religion in society. Many Americans reject the idea that religion causes more problems than it solves. Moreover, most religious individuals don’t see a conflict between their devoutness and living in a modern society.

Religion and Society

Religion and Societal Problems

Quesion: Religion causes more problems in society than it solves.

• Strong Agreement: Atheists and agnostics tend to strongly agree with the notion that religion causes more harm than good. This perspective likely stems from a belief in secularism and skepticism toward religious institutions’ influence on societal affairs.

• Varied Perspectives: Other religious groups exhibit more diverse views. For example, Mormons and Protestants largely disagree with the statement, indicating a belief in the positive contributions of religion to society. Conversely, Buddhists, Hindus, and Jews show mixed sentiments, with a significant portion expressing agreement or uncertainty.

• Neutral or Uncertain Stance: Some respondents, particularly those who volunteered as “Don’t know/Refused,” refrain from taking a definitive position on the matter. This could indicate a lack of clarity or hesitation in expressing a viewpoint.

Anti_religious_belief <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q39c)) %>%  # Ensure q16 and q39c are not NA
  group_by(q16, q39c) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q39c)


Anti_religious_belief_selected <- select(Anti_religious_belief, q16, q39c, Percentage)


Anti_religious_belief_wide <- pivot_wider(
  Anti_religious_belief_selected,
  names_from = q39c,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Anti_religious_belief_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Anti_religious_belief_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Question: Religion causes more problems in society than it solves" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Question: Religion causes more problems in society than it solves
Religion	Completely agree	Completely disagree	Don’t know/refused (VOL.)	Mostly agree	Mostly disagree
Agnostic (not sure if there is a God)	40.193705	6.174334	2.058111	32.687651	18.88620
Atheist (do not believe in God)	52.427185	5.048544	2.912621	26.990291	12.62136
Buddhist	30.900243	9.002433	6.082725	29.683698	24.33090
Christian (VOL)	15.239726	32.191781	7.705480	18.835616	26.02740
Don’t know/Refused (VOL)	27.200000	11.200000	23.600000	22.400000	15.60000
Hindu	26.848249	15.175097	3.501946	32.684825	21.78988
Jehovah’s Witness (VOL)	31.924883	22.535211	8.450704	23.474178	13.61502
Jewish (Judaism)	22.873900	15.395894	5.718475	28.445748	27.56598
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	3.270224	55.421687	1.893287	9.122203	30.29260
Muslim (Islam)	10.344828	46.551724	3.448276	18.103448	21.55172
Nothing in particular	26.975991	15.376315	5.179390	28.675479	23.79282
Orthodox (Greek, Russian, or some other orthodox church)	13.498623	29.476584	4.132231	21.763085	31.12948
Protestant	8.189397	43.736719	3.857680	15.872065	28.34414
Roman Catholic	11.494967	33.702001	3.541693	19.659500	31.60184
Unitarian (Universalist) (VOL)	27.922078	6.493506	4.545454	33.116883	27.92208

Conclusion:

The data reflects the complexity of attitudes toward religion’s societal impact, highlighting contrasting perspectives among different religious affiliations. While some groups strongly believe in religion’s detrimental effects, others emphasize its positive influence. Understanding these varied viewpoints is essential for fostering dialogue and addressing societal challenges related to religion’s role in contemporary life.

Exploring Conflicts in Modern Living

Peaceful_living <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q46)) %>%  # Ensure q16 and q41 are not NA
  group_by(q16, q46) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q46)


Peaceful_living_selected <- select(Peaceful_living, q16, q46, Percentage)


Peaceful_living_wide <- pivot_wider(
  Peaceful_living_selected,
  names_from = q46,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Peaceful_living_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Peaceful_living_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "there is natural conflict between being a devout religious person and living in a modern society," = 2, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	there is natural conflict between being a devout religious person and living in a modern society,
Religion	Don’t know/refused (VOL)	No, don’t think so	Yes, there is conflict
Buddhist	4.866180	69.34307	25.79075
Christian (VOL)	7.534247	54.96575	37.50000
Hindu	7.003891	62.64591	30.35019
Jehovah’s Witness (VOL)	9.859155	30.51643	59.62441
Jewish (Judaism)	6.011730	64.22287	29.76540
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	3.442341	61.61790	34.93976
Muslim (Islam)	5.172414	59.48276	35.34483
Orthodox (Greek, Russian, or some other orthodox church)	6.611570	60.60606	32.78237
Protestant	6.276903	51.89887	41.82422
Roman Catholic	5.231763	61.74972	33.01852
Unitarian (Universalist) (VOL)	4.545454	72.72727	22.72727

Among various religious groups, there’s a range of perspectives: The majority of Buddhists (69.3%), Hindus (62.6%), and Jews (64.2%) don’t perceive a conflict. Conversely, Jehovah’s Witnesses (59.6%), Protestants (41.8%), and Christians in general (37.5%) are more likely to see a conflict. Muslims (35.3%) and Orthodox Christians (32.8%) also lean towards perceiving a conflict. Overall, while there’s diversity in opinions, around a third to a half of respondents across different religious affiliations see a natural conflict between devout religious practice and living in a modern society.

Peaceful_living_non <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q47)) %>%  # Ensure q16 and q41 are not NA
  group_by(q16, q47) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q47)


Peaceful_living_non_selected <- select(Peaceful_living_non, q16, q47, Percentage)


Peaceful_living_non_wide <- pivot_wider(
  Peaceful_living_non_selected,
  names_from = q47,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Peaceful_living_non_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Peaceful_living_non_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "there is a natural conflict between being a non-religious person and living in a society where most people are religious," = 2, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	there is a natural conflict between being a non-religious person and living in a society where most people are religious,
Religion	Don’t know/refused (VOL)	No, don’t think so	Yes, there is conflict
Agnostic (not sure if there is a God)	4.358353	57.26392	38.37772
Atheist (do not believe in God)	4.660194	52.23301	43.10680
Don’t know/Refused (VOL)	27.600000	51.20000	21.20000
Nothing in particular	6.123550	64.31076	29.56569

Agnostics (38.4%) and atheists (43.1%) are more likely to perceive a conflict. Those with no religious affiliation, such as “nothing in particular,” also show a similar trend, with 29.6% perceiving a conflict. However, a majority of respondents in these categories (57.3% for agnostics, 52.2% for atheists, and 64.3% for those with no particular affiliation) don’t perceive such a conflict. Notably, a significant proportion of respondents who refused to answer or didn’t know expressed uncertainty (27.6%) or didn’t perceive a conflict (51.2%). Overall, while a notable portion of non-religious individuals perceive a conflict between their beliefs and a predominantly religious society, the majority across these categories do not share this perception.

Satisfaction

First I want to go through all quesions about the different aspects of satisfaction.

Quesion:Are you very satisfied, somewhat satisfied, somewhat DISsatisfied or very DISsatisfied? (Your standard of living – what you can buy or do)

Personal_satisfaction <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q3a)) %>%  # Ensure q16 and q3a are not NA
  group_by(q16, q3a) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q3a)


Personal_satisfaction_selected <- select(Personal_satisfaction, q16, q3a, Percentage)


Personal_satisfaction_wide <- pivot_wider(
  Personal_satisfaction_selected,
  names_from = q3a,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Personal_satisfaction_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Personal_satisfaction_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Personal Life Satisfaction" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

table

	Personal Life Satisfaction
Religion	Don’t know/Refused (VOL.)	Somewhat dissatisfied	Somewhat satisfied	Very dissatisfied	Very satisfied
Agnostic (not sure if there is a God)	0.9685230	11.864407	38.37772	5.690073	43.09927
Atheist (do not believe in God)	1.3592233	9.514563	39.02913	6.796117	43.30097
Buddhist	0.7299270	10.948905	41.36253	5.109489	41.84915
Christian (VOL)	2.5684932	11.986301	41.43836	9.246575	34.76027
Don’t know/Refused (VOL)	4.0000000	12.400000	37.20000	6.000000	40.40000
Hindu	0.0000000	4.669261	43.96887	5.058366	46.30350
Jehovah’s Witness (VOL)	2.8169014	13.615023	40.84507	10.328639	32.39437
Jewish (Judaism)	1.1730205	8.944282	36.21701	4.252199	49.41349
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	0.5163511	10.327022	36.14458	4.647160	48.36489
Muslim (Islam)	0.8620690	8.620690	44.82759	7.758621	37.93103
Nothing in particular	1.2139196	13.002428	41.40815	8.740221	35.63528
Orthodox (Greek, Russian, or some other orthodox church)	2.7548209	12.947658	40.77135	6.060606	37.46556
Protestant	1.2423037	11.464066	38.34251	7.181387	41.76974
Roman Catholic	1.3669691	11.905058	39.67938	6.648440	40.40015
Unitarian (Universalist) (VOL)	1.9480519	9.090909	38.31169	3.896104	46.75325

Quesion: Are you very satisfied, somewhat satisfied, somewhat DISsatisfied or very DISsatisfied? (With your family life)

family_life_satisfaction <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q3b)) %>%  # Ensure q16 and q3a are not NA
  group_by(q16, q3b) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q3b)


family_life_satisfaction_selected <- select(family_life_satisfaction, q16, q3b, Percentage)


family_life_satisfaction_wide <- pivot_wider(
  family_life_satisfaction_selected,
  names_from = q3b,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(family_life_satisfaction_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(family_life_satisfaction_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Family Life Satisfaction" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

table

	Family Life Satisfaction
Religion	Don’t know/Refused (VOL.)	Somewhat dissatisfied	Somewhat satisfied	Very dissatisfied	Very satisfied
Agnostic (not sure if there is a God)	1.5738499	3.389830	22.39709	1.8159806	70.82324
Atheist (do not believe in God)	4.2718447	2.524272	22.33010	1.9417476	68.93204
Buddhist	1.2165450	2.433090	22.14112	1.4598540	72.74939
Christian (VOL)	4.1095890	3.424657	18.15068	2.3972603	71.91781
Don’t know/Refused (VOL)	8.0000000	2.400000	22.00000	2.4000000	65.20000
Hindu	1.9455253	2.723735	17.12062	1.5564202	76.65370
Jehovah’s Witness (VOL)	1.8779343	6.572770	21.12676	1.8779343	68.54460
Jewish (Judaism)	2.7859238	3.519062	16.42229	1.9061584	75.36657
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	0.8605852	2.925990	13.42513	0.3442341	82.44406
Muslim (Islam)	0.8620690	2.586207	20.68966	3.4482759	72.41379
Nothing in particular	2.4008632	3.345023	20.36687	2.3738872	71.51335
Orthodox (Greek, Russian, or some other orthodox church)	2.4793388	3.856749	18.73278	3.0303030	71.90083
Protestant	1.9996731	3.100311	17.29962	1.7326868	75.86771
Roman Catholic	1.5658009	3.193737	15.55859	1.5906549	78.09121
Unitarian (Universalist) (VOL)	3.8961039	0.000000	12.98701	0.6493506	82.46753

Quesion:Are you very satisfied, somewhat satisfied, somewhat DISsatisfied or very DISsatisfied? (With the way the political system is working in this country)

Standard_of_living_satisfaction <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q3c)) %>%  # Ensure q16 and q3a are not NA
  group_by(q16, q3c) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q3c)


Standard_of_living_satisfaction_selected <- select(Standard_of_living_satisfaction, q16, q3c, Percentage)


Standard_of_living_satisfaction_wide <- pivot_wider(
  Standard_of_living_satisfaction_selected,
  names_from = q3c,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Standard_of_living_satisfaction_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Standard_of_living_satisfaction_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Standard of Life Satisfaction" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

table

	Standard of Life Satisfaction
Religion	Don’t know/Refused (VOL.)	Somewhat dissatisfied	Somewhat satisfied	Very dissatisfied	Very satisfied
Agnostic (not sure if there is a God)	1.937046	26.99758	12.590799	56.41646	2.058111
Atheist (do not believe in God)	1.553398	22.13592	13.980583	57.86408	4.466019
Buddhist	2.919708	17.27494	15.328467	59.85401	4.622871
Christian (VOL)	5.821918	25.17123	17.808219	46.57534	4.623288
Don’t know/Refused (VOL)	14.800000	21.20000	12.400000	45.60000	6.000000
Hindu	6.614786	24.90272	24.513619	34.63035	9.338521
Jehovah’s Witness (VOL)	28.169014	16.43192	7.042254	46.47887	1.877934
Jewish (Judaism)	2.199413	24.78006	19.648094	47.50733	5.865103
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	3.098107	28.22719	26.678141	34.42341	7.573150
Muslim (Islam)	5.172414	17.24138	15.517241	52.58621	9.482759
Nothing in particular	4.585919	23.73887	16.994875	50.41813	4.262207
Orthodox (Greek, Russian, or some other orthodox church)	6.336088	26.17080	24.517906	35.26171	7.713499
Protestant	3.536207	25.90312	21.402496	43.96012	5.198060
Roman Catholic	3.641109	25.83572	21.809370	42.51274	6.201069
Unitarian (Universalist) (VOL)	1.298701	19.48052	5.844156	72.07792	1.298701

Quesion: Are you very satisfied, somewhat satisfied, somewhat DISsatisfied or very DISsatisfied? (With your personal safety from things like crime and terrorism)

Personal_Safety <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q3d)) %>%  # Ensure q16 and q3a are not NA
  group_by(q16, q3d) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q3d)


Personal_Safety_selected <- select(Personal_Safety, q16, q3d, Percentage)


Personal_Safety_wide <- pivot_wider(
  Personal_Safety_selected,
  names_from = q3d,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Personal_Safety_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Personal_Safety_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Personal Safety Satisfaction" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

table

	Personal Safety Satisfaction
Religion	Don’t know/Refused (VOL.)	Somewhat dissatisfied	Somewhat satisfied	Very dissatisfied	Very satisfied
Agnostic (not sure if there is a God)	1.937046	10.65375	42.25182	4.721550	40.43584
Atheist (do not believe in God)	2.524272	12.81553	36.69903	5.631068	42.33010
Buddhist	4.379562	11.43552	39.41606	5.839416	38.92944
Christian (VOL)	3.938356	14.72603	37.50000	10.445206	33.39041
Don’t know/Refused (VOL)	7.600000	16.00000	35.20000	10.800000	30.40000
Hindu	4.280156	11.28405	34.63035	12.840467	36.96498
Jehovah’s Witness (VOL)	6.103286	16.90141	30.04695	29.107981	17.84038
Jewish (Judaism)	2.346041	14.22287	44.28152	7.038123	32.11144
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	1.549053	12.04819	36.66093	6.368330	43.37349
Muslim (Islam)	2.586207	13.79310	36.20690	10.344828	37.06897
Nothing in particular	3.129215	12.51686	40.94955	9.738333	33.66604
Orthodox (Greek, Russian, or some other orthodox church)	5.234160	11.84573	39.39394	13.498623	30.02755
Protestant	2.473710	14.19386	38.91462	10.690350	33.72746
Roman Catholic	2.696657	14.82540	39.59239	11.445259	31.44029
Unitarian (Universalist) (VOL)	3.896104	10.38961	36.36364	5.844156	43.50649

Now based on this quesions and answer I’m gonna create the measure of Satisfaction within all individuals. Certainly! Here’s the explanation you can use:

Total Satisfaction Metric Explanation:

To quantify overall satisfaction from responses to four questions (q3a, q3b, q3c, q3d) in the survey, a Total Satisfaction Metric is computed. This metric converts qualitative responses into numerical scores using the following mapping:

• ‘Very satisfied’: Scored as 1
• ‘Somewhat satisfied’: Scored as 0.75
• ‘Somewhat dissatisfied’: Scored as 0.25
• ‘Very dissatisfied’: Scored as 0
• ‘Don’t know/Refused’: Scored as 0.5 ( My idea was that if somebody can’t classify his answers to binary class of dissatisfied or satisfied, he/she probably has higher satisfaction than someone with dissatisfied answers)

These scores represent the level of satisfaction associated with each response, with higher scores indicating higher levels of satisfaction.

The Total Satisfaction Metric is calculated by summing the scores from the four questions for each respondent. The resulting score provides an aggregate measure of satisfaction, allowing for easy comparison across respondents.

Mathematically, the Total Satisfaction Metric (\(TSM\)) can be represented as:

\[ TSM = \sum_{i=1}^{4} Score_i \]

Where: - \(TSM\) is the Total Satisfaction Metric, - \(Score_i\) is the score assigned to the response for the \(i^{th}\) question.

map_satisfaction_scores <- function(satisfaction, detail) {
  case_when(
    satisfaction == 'Satisfied' & detail == 'Very' ~ 1,
    satisfaction == 'Satisfied' & detail == 'Somewhat' ~ 0.75,
    satisfaction == 'Dissatisfied' & detail == 'Very' ~ 0,
    satisfaction == 'Dissatisfied' & detail == 'Somewhat' ~ 0.25,
    # Assuming 'Don’t know/Refused' is for cases where detail might not be answered
    satisfaction == 'Satisfied' & detail == 'Don’t know/Refused' ~ 0.75, 
    satisfaction == 'Dissatisfied' & detail == 'Don’t know/Refused' ~ 0.25,
    TRUE ~ 0.5  # Default case for any other combinations, including unhandled 'Don’t know/Refused'
  )
}

# Apply this mapping function to your dataset for Q1/Q1a and Q2/Q2a
survey_data_2007 <- survey_data_2007 %>%
  mutate(q1_Score = map_satisfaction_scores(as.character(q1), as.character(q1a)),
         q2_Score = map_satisfaction_scores(as.character(q2), as.character(q2a)))


# Map scores for Q3 questions directly
satisfaction_mapping <- c('Very satisfied' = 1, 'Somewhat satisfied' = 0.75, 'Somewhat dissatisfied' = 0.25, 'Very dissatisfied' = 0, 'Don’t know/Refused' = 0.5)

survey_data_2007 <- survey_data_2007 %>%
  mutate(across(c(q3a, q3b, q3c, q3d), ~ satisfaction_mapping[.], .names = "{.col}_Score"))

# Calculate total satisfaction score
survey_data_2007$Total_Satisfaction <- rowSums(survey_data_2007[, grepl("_Score$", names(survey_data_2007))], na.rm = TRUE)

# Viewing the results
head(survey_data_2007[, c("q1_Score", "q2_Score", "q3a_Score", "q3b_Score", "q3c_Score", "q3d_Score", "Total_Satisfaction")])

average_satisfaction_by_religion <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>%
  summarise(Average_Satisfaction = mean(Total_Satisfaction, na.rm = TRUE))


# Plot the data using ggplot
p <- ggplot(average_satisfaction_by_religion, aes(x = q16, y = Average_Satisfaction, fill = q16)) +
  geom_bar(stat = "identity", width = 0.7) +  # Use identity to indicate that y values are already calculated
  
  labs(title = "Average Satisfaction by Religion",
       x = "Religion",
       y = "Average Satisfaction") +
  
  theme_minimal() +  # Use a minimal theme
  theme(axis.text.x = element_text(angle = 45, hjust = 1),  # Adjust text alignment if needed
        legend.position = "none")  # Remove legend since color coding is redundant

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 800)

# Show the interactive plot in your default browser

interactive_plot

Average Satisfaction by Religious Affiliation:

The average satisfaction levels for various religious affiliations provide valuable insights into the overall contentment reported by individuals within each group. While there are fluctuations across different affiliations, the average satisfaction score serves as a broad indicator of the perceived fulfillment among members of each religious community.

Across the spectrum, the average satisfaction tends to hover around a certain range, typically between 2.4 and 2.7. However, some religious affiliations stand out with notably higher or lower average satisfaction scores.

For instance, Hindus display a relatively high average satisfaction score of 2.70, indicating a generally contented outlook within this religious community. On the other hand, Jehovah’s Witnesses report a lower average satisfaction score of 2.08, suggesting a lower overall satisfaction level among its members.

Among the affiliations with average satisfaction scores that deviate from the norm, Mormons, with an average satisfaction score of 2.80, demonstrate a notably higher level of contentment compared to the general trend. Conversely, individuals identifying as Don’t know/Refused exhibit a slightly lower average satisfaction score of 2.35, indicating a relatively lower reported satisfaction level within this group.

# Create a density plot to visualize the distribution of Total_Satisfaction
p <- ggplot(survey_data_2007, aes(x = Total_Satisfaction)) +
  geom_density(fill = "lightblue", adjust = 4) +
  labs(title = "Density Plot of Total Satisfaction Scores",
       x = "Total Satisfaction Score",
       y = "Density") +
  theme_minimal()

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 500)

# Show the interactive plot in your default browser

interactive_plot

Happiness

Based on the data provided, it seems that individuals who identify with a religious affiliation generally report higher levels of happiness compared to those who do not identify with any religion. Across various religious groups, there appears to be a trend of higher reported happiness levels, with a significant portion of respondents indicating that they are either “Pretty happy” or “Very happy.”

Interestingly, even within religious groups, there is variation in reported happiness levels. For example, Mormons and Unitarian Universalists tend to report particularly high levels of happiness, while Jehovah’s Witnesses and Christians who volunteered their religion report lower levels of happiness compared to some other religious groups.

Overall, while there are differences among various religious affiliations, the data suggests a positive association between religious identification and reported happiness levels.

Happiness <- survey_data_2014 %>%
  group_by(qe1) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(qe1), !is.na(qa1)) %>%  # Ensure qe1 and qa1 are not NA
  group_by(qe1, qa1) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(qe1) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(qe1, qa1)
Happiness_selected <- select(Happiness, qe1, qa1, Percentage)


Happiness_wide <- pivot_wider(
  Happiness_selected,
  names_from = qa1,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Happiness_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Happiness_wide, "html", align = c('l', rep('c', 10))) %>%
  kable_styling("striped", full_width = F) %>%
  add_header_above(c(" " = 1, "How do you describe your Happiness state" = 3, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	How do you describe your Happiness state
Religion	(VOL) Don’t know/Refused	Not too happy	Pretty happy	Very happy
Agnostic (not sure if there is a God)	1.289009	10.379919	60.44776	27.88331
Atheist (do not believe in God)	1.548270	9.107468	59.10747	30.23679
Buddhist	2.651515	9.469697	57.95455	29.92424
Christian (VOL)	3.466667	15.600000	48.53333	32.40000
Don’t Know/Refused (VOL)	7.792208	5.194805	51.94805	35.06494
Hindu	1.507538	9.547739	59.29648	29.64824
Jehovah’s Witness (VOL)	4.098361	19.672131	42.21311	34.01639
Jewish (Judaism)	1.534829	11.570248	52.77450	34.12042
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	1.355422	9.186747	51.35542	38.10241
Muslim (Islam)	2.109705	13.080169	50.63291	34.17722
Nothing in particular	1.845907	12.961477	54.35393	30.83868
Orthodox (Greek, Russian, or some other orthodox church)	4.838710	11.290323	54.83871	29.03226
Protestant	2.032572	12.643606	50.50499	34.81884
Roman Catholic	2.543080	13.896609	49.51362	34.04669
Spiritual but not religious (‘believe in God, but…’) (VOL)	6.250000	13.281250	46.87500	33.59375
Unitarian (Universalist) (VOL)	1.212121	5.454546	56.36364	36.96970

p <- ggplot(Happiness, aes(x = qe1, y = qa1, fill = Percentage)) +
  geom_tile() + # Create the tiles for the heatmap
  scale_fill_gradient(low = "white", high = "steelblue") + # Gradient from white (low) to steel blue (high)
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1), axis.text.y = element_text(size = 12), title = element_text(size = 20)) +
  labs(title = "Happiness among different Religion",
       x = "Religion",
       y = "Happiness",
       fill = "Percentage")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 750)

interactive_plot <- interactive_plot %>% layout(autosize = F)
# Show the interactive plot in your default browser

interactive_plot

Average Salary

Analyzing the average income across different religious affiliations provides valuable insights into the economic status and financial well-being of individuals within each group. These figures offer a broad overview of the income levels reported by members of various religious communities, highlighting potential disparities and trends.

# Map income categories to approximate average values
income_mapping <- c(
    'Less than $10,000' = 5000,
    '10 to under $20,000' = 15000,
    '20 to under $30,000' = 25000,
    '30 to under $40,000' = 35000,
    '40 to under $50,000' = 45000,
    '50 to under $75,000' = 62500,
    '75 to under $100,000' = 87500,
    '100 to under $150,000' = 125000,
    '$150,000 or more' = 200000,
    "Don't know/Refused (VOL.)" = NA  # Handle unknown/refused with NA or an appropriate value
)

# Assuming survey_data_2007 is your dataframe
survey_data_2007 <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n()>100) %>% 
  mutate(Average_Income = income_mapping[income])

average_income_by_religion <- survey_data_2007 %>%
  group_by(q16) %>%
  summarise(Average_Income = mean(Average_Income, na.rm = TRUE))  # Excluding NA values

sorted_average_income_by_religion <- average_income_by_religion %>%
  arrange(Average_Income)

# Ensure 'q16' is ordered according to the sorted data
sorted_average_income_by_religion$q16 <- factor(sorted_average_income_by_religion$q16, 
                                                levels = sorted_average_income_by_religion$q16)

# Plot the data using ggplot
p <- ggplot(sorted_average_income_by_religion, aes(x = q16, y = Average_Income, fill = q16)) +
  geom_bar(stat = "identity", width = 0.7) +  # Use identity to indicate that y values are already calculated
  
  labs(title = "Average Income by Religion",
       x = "Religion",
       y = "Average Satisfaction") +
  
  theme_minimal() +  # Use a minimal theme
  theme(axis.text.x = element_text(angle = 45, hjust = 1),  # Adjust text alignment if needed
        legend.position = "none")  # Remove legend since color coding is redundant

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 800)

# Show the interactive plot in your default browser

interactive_plot

The average income among different religious affiliations varies significantly, reflecting diverse economic circumstances and opportunities experienced by individuals within each group.

For instance, Hindus and Jews report relatively higher average incomes compared to other religious affiliations, with Hindus averaging $109,829.55 and Jews averaging $108,230.77. These figures suggest a relatively affluent economic status among members of these religious communities.

On the other hand, Jehovah’s Witnesses report a notably lower average income of $48,531.07, indicating a comparatively lower economic standing within this religious group.

Among the affiliations with average income figures deviating from the norm, atheists and agnostics display relatively higher average incomes, with atheists averaging $85,928.25 and agnostics averaging $79,537.67. Conversely, individuals identifying as Don’t know/Refused report a slightly lower average income of $72,870.37, indicating a relatively lower reported income level within this group.

# Create a density plot to visualize the distribution of Total_Satisfaction
p <- ggplot(survey_data_2007, aes(x = Average_Income)) +
  geom_density(fill = "lightblue", adjust = 4) +
  labs(title = "Density Plot of Income",
       x = "Total Satisfaction Score",
       y = "Density") +
  theme_minimal()

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 500)

## Warning: Removed 6047 rows containing non-finite outside the scale range
## (`stat_density()`).

# Show the interactive plot in your default browser

interactive_plot

Religion and “Culture War” Issues

Abortion

Abortion is a deeply complex and contentious issue that has sparked significant debate and controversy across various societies and cultures. Defined as the termination of a pregnancy by removing an embryo or fetus from the uterus before it can survive outside the womb, abortion raises profound ethical, moral, religious, and legal questions.

In the United States, discussions surrounding abortion have been central to public discourse and policymaking for decades. The fundamental question of whether abortion should be legally permissible and under what circumstances has divided individuals, communities, and religious groups, reflecting differing perspectives on the sanctity of life, bodily autonomy, women’s rights, and the role of government in regulating reproductive healthcare.

Against this backdrop, understanding the views of individuals within different religious affiliations regarding abortion provides valuable insights into the diverse range of beliefs, values, and moral considerations shaping public opinion on this complex issue.

Abortion <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q7)) %>%  # Ensure q16 and q7 are not NA
  group_by(q16, q7) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q7)


Abortion_selected <- select(Abortion, q16, q7, Percentage)


Abortion_wide <- pivot_wider(
  Abortion_selected,
  names_from = q7,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Abortion_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Abortion_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Abortion Should Be…" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Abortion Should Be…
Religion	(DO NOT READ) Don’t Know/Refused	Illegal in all cases	Illegal in most cases	Legal in all cases	Legal in most cases
Agnostic (not sure if there is a God)	3.389830	1.331719	10.774818	34.987893	49.51574
Atheist (do not believe in God)	4.271845	4.077670	6.990291	44.466019	40.19417
Buddhist	6.082725	2.433090	6.326034	37.712895	47.44526
Christian (VOL)	11.472603	13.184931	25.513699	14.554794	35.27397
Don’t know/Refused (VOL)	21.600000	6.000000	8.000000	26.800000	37.60000
Hindu	7.003891	4.280156	20.233463	21.789883	46.69261
Jehovah’s Witness (VOL)	7.981221	55.399061	21.126761	5.164319	10.32864
Jewish (Judaism)	2.492669	4.252199	6.744868	43.255132	43.25513
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	3.958692	8.950086	60.757315	6.540447	19.79346
Muslim (Islam)	4.310345	12.068965	33.620690	14.655172	35.34483
Nothing in particular	6.123550	7.445374	16.859995	28.324791	41.24629
Orthodox (Greek, Russian, or some other orthodox church)	8.264463	7.988981	18.732782	24.517906	40.49587
Protestant	6.293249	16.765651	31.035798	13.834250	32.07105
Roman Catholic	6.872126	17.211383	27.824034	15.210638	32.88182
Unitarian (Universalist) (VOL)	5.194805	1.298701	3.246753	46.753247	43.50649

Notably, there is considerable diversity in attitudes towards abortion among different religious affiliations:

• Christian Denominations: Within Christian denominations, opinions vary widely, with some groups, such as Jehovah’s Witnesses and Mormons, expressing stronger opposition to abortion, while others, like Unitarians and Protestants, exhibit more nuanced views, including significant support for abortion rights.

• Non-Christian Religions: Hindus, Muslims, and Buddhists also display a range of perspectives, with differing proportions advocating for different levels of legality regarding abortion.

• Non-Affiliated Individuals: In contrast, individuals identifying as having “Nothing in particular” or providing a “Don’t Know/Refused” response tend to demonstrate more nuanced views on abortion, with a greater willingness to support legal access to abortion in certain circumstances.

Homosexuality

The topic of homosexuality elicits diverse perspectives, often influenced by cultural, religious, and personal beliefs. Understanding attitudes towards homosexuality within different religious communities provides valuable insights into the intersections of faith and social acceptance.

Homosexuality <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q5b)) %>%  # Ensure q16 and q7 are not NA
  group_by(q16, q5b) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q5b)


Homosexuality_selected <- select(Homosexuality, q16, q5b, Percentage)


Homosexuality_wide <- pivot_wider(
  Homosexuality_selected,
  names_from = q5b,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Homosexuality_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Homosexuality_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Homosexuality Should Be…" = 3, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Homosexuality Should Be…
Religion	DK/REF (VOL)	Homosexuality is a way of life that should be accepted by so	Homosexuality is a way of life that should be discouraged by	Neither/Both equally (VOL)
Agnostic (not sure if there is a God)	3.026634	81.96126	10.048426	4.963680
Atheist (do not believe in God)	2.912621	81.94175	12.427184	2.718447
Buddhist	3.649635	85.15815	8.759124	2.433090
Christian (VOL)	8.390411	45.71918	39.212329	6.678082
Don’t know/Refused (VOL)	13.200000	68.00000	14.000000	4.800000
Hindu	10.894942	48.24903	36.964981	3.891051
Jehovah’s Witness (VOL)	5.164319	10.79812	77.464789	6.572770
Jewish (Judaism)	3.079179	80.20528	13.343108	3.372434
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	3.958692	21.17040	69.879518	4.991394
Muslim (Islam)	3.448276	35.34483	50.862069	10.344828
Nothing in particular	5.071486	67.97950	22.255193	4.693822
Orthodox (Greek, Russian, or some other orthodox church)	9.090909	47.10744	36.639119	7.162534
Protestant	4.958317	37.59603	52.018743	5.426906
Roman Catholic	6.225923	56.38126	31.676401	5.716416
Unitarian (Universalist) (VOL)	1.298701	92.20779	2.597403	3.896104

• Acceptance vs. Disapproval: Individuals from religious backgrounds, such as Jehovah’s Witnesses, Mormons, and Muslims, tend to express more disapproval towards homosexuality. This aligns with teachings in these faiths that may view homosexuality as inconsistent with religious doctrine or moral principles.

• Moderate Views: Certain religious groups, such as Protestants and Orthodox Christians, exhibit more moderate perspectives on homosexuality, with a sizable portion expressing neither strong acceptance nor strong disapproval. This suggests a degree of diversity within these religious communities regarding attitudes towards homosexuality.

• Acceptance: Conversely, individuals identifying as agnostic, atheist, or having “Nothing in particular” as their religious affiliation tend to demonstrate higher levels of acceptance towards homosexuality. This reflects a trend observed among those who may not adhere to specific religious teachings or who prioritize secular values over religious doctrines.

The data underscores the complexity of attitudes towards homosexuality within different religious contexts. While some religious groups exhibit more conservative views, others demonstrate greater acceptance or neutrality. These findings highlight the multifaceted relationship between religion and social attitudes, emphasizing the need for dialogue and understanding across diverse religious perspectives in discussions surrounding LGBTQ+ rights and social inclusion.

Evolution

Question: Evolution is the best explanation for the origins of human life on earth
Debates over the origins of human life have long been a focal point in discussions on science, religion, and philosophy. One of the key areas of contention is the acceptance of evolution as the primary explanation for the origins of human life. Understanding the perspectives of different religious groups on this issue provides valuable insights into the intersection of faith and scientific understanding.

Evolution <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q10c)) %>%  # Ensure q16 and q7 are not NA
  group_by(q16, q10c) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q10c)


Evolution_selected <- select(Evolution, q16, q10c, Percentage)


Evolution_wide <- pivot_wider(
  Evolution_selected,
  names_from = q10c,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Evolution_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Evolution_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Evolution as the Best Explanation for the Origins of Human Life" = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Evolution as the Best Explanation for the Origins of Human Life
Religion	Completely agree	Completely disagree	Don’t know/Refused (VOL.)	Mostly agree	Mostly disagree
Agnostic (not sure if there is a God)	57.142857	3.753027	2.663438	30.629540	5.811138
Atheist (do not believe in God)	71.067961	4.271845	2.718447	17.087379	4.854369
Buddhist	51.581508	4.136253	4.866180	31.386861	8.029197
Christian (VOL)	16.095890	32.705480	10.102740	22.260274	18.835616
Don’t know/Refused (VOL)	38.800000	11.200000	20.000000	24.800000	5.200000
Hindu	44.357977	4.669261	4.669261	37.743191	8.560311
Jehovah’s Witness (VOL)	1.877934	81.220657	1.877934	3.755869	11.267606
Jewish (Judaism)	51.026393	8.944282	4.985337	28.592375	6.451613
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	5.851979	55.593804	2.581756	14.629948	21.342513
Muslim (Islam)	17.241379	36.206897	5.172414	27.586207	13.793103
Nothing in particular	33.072565	12.408956	7.202590	34.637173	12.678716
Orthodox (Greek, Russian, or some other orthodox church)	22.314050	20.661157	8.815427	33.333333	14.876033
Protestant	11.823680	40.887048	6.750940	23.156977	17.381355
Roman Catholic	21.262582	18.317385	6.983969	33.813844	19.622219
Unitarian (Universalist) (VOL)	62.337662	3.246753	2.597403	26.623377	5.194805

• High Acceptance: Individuals identifying as atheists, agnostics, and Unitarians demonstrate the highest levels of agreement with evolution as the best explanation for human life’s origins. This aligns with the scientific consensus and reflects a tendency among those with more secular or progressive religious views to embrace evolutionary theory.

• Moderate Acceptance: Certain religious groups, such as Buddhists and Jews, exhibit moderate levels of agreement with evolution. While not as high as atheists or agnostics, these groups still show a substantial proportion endorsing evolutionary theory, suggesting a degree of compatibility between their religious beliefs and scientific understanding.

• Low Acceptance: Conversely, individuals from more conservative religious backgrounds, such as Jehovah’s Witnesses, Mormons, and Muslims, tend to express lower levels of agreement with evolution. This likely reflects theological teachings or interpretations that may conflict with evolutionary principles.

Conclusion: The data highlights the diversity of perspectives among different religious groups regarding the acceptance of evolution as the best explanation for human life’s origins. While some religious traditions demonstrate high levels of acceptance, others exhibit more skepticism or outright rejection of evolutionary theory. These findings underscore the complex relationship between religion and science, emphasizing the importance of fostering dialogue and understanding across diverse religious perspectives in discussions concerning the origins of human life.

Churches and Politics

Question: In your opinion, should churches and other houses of worship keep out of political matters - or should they express their views on day-to-day social and political questions?

The role of churches in expressing views on political matters is a topic of considerable debate and contention. While some argue that churches should actively engage in political discourse as part of their societal role, others advocate for a separation of church and state, suggesting that religious institutions should refrain from direct involvement in political affairs.

Church_politics <- survey_data_2007 %>%
  group_by(q16) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(q16), !is.na(q9)) %>%  # Ensure q16 and q7 are not NA
  group_by(q16, q9) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(q16) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(q16, q9)


Church_politics_selected <- select(Church_politics, q16, q9, Percentage)


Church_politics_wide <- pivot_wider(
  Church_politics_selected,
  names_from = q9,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Church_politics_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Church_politics_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "Should Churches Express Views on Politics?" = 2, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Should Churches Express Views on Politics?
Religion	Don’t know/Refused	Should express views	Should keep out
Agnostic (not sure if there is a God)	3.510896	27.845036	68.64407
Atheist (do not believe in God)	4.077670	22.524272	73.39806
Buddhist	7.785888	34.063260	58.15085
Christian (VOL)	5.821918	48.801370	45.37671
Don’t know/Refused (VOL)	10.000000	29.200000	60.80000
Hindu	6.225681	22.178988	71.59533
Jehovah’s Witness (VOL)	6.572770	9.859155	83.56808
Jewish (Judaism)	3.958944	31.818182	64.22287
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	4.991394	44.922547	50.08606
Muslim (Islam)	0.862069	52.586207	46.55172
Nothing in particular	4.882654	34.367413	60.74993
Orthodox (Greek, Russian, or some other orthodox church)	4.132231	39.669421	56.19835
Protestant	4.211845	56.568408	39.21975
Roman Catholic	4.200323	46.737915	49.06176
Unitarian (Universalist) (VOL)	7.142857	35.714286	57.14286

• Support for Expression: Individuals identifying as Protestants, Mormons, and Christians generally favor churches expressing views on political matters. This inclination may stem from a belief in the moral authority of religious institutions and their role in shaping societal values and policies.

• Preference for Neutrality: Conversely, atheists, agnostics, and Unitarians tend to advocate for churches keeping out of political issues. This reflects a desire to maintain a clear separation between religious and political domains, preserving individual autonomy and diversity of belief.

• Moderate Views: Some religious groups, such as Buddhists, Jews, and Hindus, demonstrate more mixed perspectives. While they may acknowledge the potential for churches to engage in political discourse, they also recognize the importance of maintaining impartiality and avoiding undue influence on political processes.

Conclusion

The survey data from various religious and non-religious groups reveal profound insights into how religion influences social perceptions, behaviors, and personal satisfaction within society. It suggests that religion can be both a source of conflict and a provider of community, shaping attitudes toward modern societal norms and personal well-being in diverse ways.

Influence on Societal Views: Individuals from religious backgrounds often view their faith as a stabilizing force, contributing positively to their lives and communities. This is evident in the higher satisfaction rates regarding family and personal life among those with strong religious affiliations like Hindus and Mormons. In contrast, non-religious individuals, including atheists and agnostics, tend to view religion more critically, often perceiving it as a source of social tension or conflict. This group is more likely to believe that religion complicates social issues rather than resolves them, highlighting a stark difference in how the role of religion in society is perceived.

Impact on Ethical and Social Issues: The data shows that religious beliefs significantly influence attitudes towards contentious social and ethical issues such as abortion and homosexuality. More conservative views are prevalent among traditionally religious groups like Jehovah’s Witnesses and Mormons, while more liberal views tend to be held by those who are non-religious or part of more progressive religious groups like Unitarians. This division underscores how religious doctrines can shape fundamental social attitudes and behaviors.

Adaptation to Modernity: There is a perceived conflict between devout religious practice and living in a modern society, particularly among groups with strict doctrinal adherence. However, many believers find ways to reconcile their faith with modern values, indicating that religion evolves with societal changes. Non-religious individuals also face their own challenges, particularly in societies where religious norms predominate, suggesting a universal theme of navigating personal beliefs in diverse social landscapes.

General Well-being and Religion: The relationship between religion and personal satisfaction is complex. While religious affiliation often correlates with higher levels of family and personal life satisfaction—possibly due to stronger community bonds and support systems—the less religious report varying degrees of satisfaction, influenced by their personal and social environments.

Chapter 3 : The Changing Religious Trends

Changes in Religious Affiliations

In this section, we will focus on analyzing the change in religious affiliations between 2007 and 2014. By examining how the proportions of each religious group shifted over time, we aim to uncover trends and patterns in religious identification within the surveyed population. This analysis will provide insights into the evolving religious landscape and shed light on the dynamics driving changes in religious adherence during the specified period.

<br

1. Overall Trends:
   • The category “Nothing in particular” experienced the most significant increase in both absolute count and percentage change, indicating a rise in individuals who don’t identify with any specific religious affiliation.
   • Protestantism witnessed a substantial decline both in count and percentage, indicating a decrease in adherence to Protestant denominations.
   • Other religious groups like Buddhism, Hinduism, and Orthodox Christianity saw declines in counts and percentages, albeit to a lesser extent compared to Protestantism.
   • Atheism and Agnosticism saw notable increases in both count and percentage, suggesting a rise in non-theistic beliefs or uncertainties about the existence of God.
   • Muslim and Jewish communities experienced moderate increases in number and percentage, indicating potential growth or stability in their populations.

# Assuming 'qe1' is the column for religion
religion_counts_2014 <- survey_data_2014 %>%
  count(qe1) %>%
  filter(n > 100)

# Rename the columns for clarity
colnames(religion_counts_2014) <- c("Religion", "Respondents")


# Calculate the total number of respondents
total_respondents <- sum(religion_counts_2014$Respondents)

# Calculate percentages
religion_counts_2014 <- religion_counts_2014 %>%
  mutate(Percentage = (Respondents / total_respondents) * 100)


# Rename the columns for clarity
colnames(religion_counts_2007) <- c("Religion", "Respondents_2007","Percentage_2007")
colnames(religion_counts_2014) <- c("Religion", "Respondents_2014","Percentage_2014")

# Assuming religion_counts_2007 and religion_counts_2014 are your data frames
# First, ensure that the 'Religion' column is treated as a character for accurate comparisons
religion_counts_2007$Religion <- as.character(religion_counts_2007$Religion)
religion_counts_2014$Religion <- as.character(religion_counts_2014$Religion)

# Identify religions present in both 2007 and 2014
common_religions <- intersect(religion_counts_2007$Religion, religion_counts_2014$Religion)

# Filter both data frames to include only common religions
religion_counts_2007 <- religion_counts_2007 %>%
  filter(Religion %in% common_religions)

religion_counts_2014 <- religion_counts_2014 %>%
  filter(Religion %in% common_religions)

# Merge the data frames
religion_changes <- merge(religion_counts_2007, religion_counts_2014, by = "Religion")

# Calculate changes
religion_changes <- religion_changes %>%
  mutate(Count_Change = Respondents_2014 - Respondents_2007,
         Percentage_Change = Percentage_2014 - Percentage_2007)

p <- ggplot(religion_changes, aes(x = reorder(Religion, -Percentage_Change), y = Percentage_Change, fill = Religion)) +
  geom_col() +
  coord_flip() +
  labs(title = "Change in Religious Affiliation from 2007 to 2014",
       x = "Religion",
       y = "Change in Count") +
  theme_minimal() +
  theme(legend.position = "none")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 500)

# Show the interactive plot in your default browser

interactive_plot

2. Specific Changes:
   • Roman Catholicism and Protestantism, the two largest religious groups in the dataset, experienced declines in both count and percentage, with Protestantism witnessing a relatively larger decrease.
   • While some religious groups saw declines in counts, their percentage changes varied due to shifts in the overall respondent population.
   • Jehovah’s Witness and Mormon denominations experienced slight increases in counts and percentages, indicating stability or minor growth within these communities.

These changes suggest a dynamic landscape of religious affiliation over the examined period, with shifts towards non-affiliation or alternative religious beliefs alongside declines in traditional denominations like Protestantism and Catholicism.

Changing Perceptions of Religion’s Importance

This section provides a comparative analysis of the importance of religion in individuals’ lives across various religious groups between the years 2007 and 2014. Through a side-by-side visualization, the shift in attitudes towards religion is depicted, showcasing the percentage of respondents within each religious category who rated religion as “Not at all important,” “Not too important,” “Somewhat important,” and “Very important.” The chart illuminates trends suggesting a declining emphasis on religion over time, with some groups demonstrating a more pronounced shift towards secularism, while others maintain a steadfast commitment to religious values. This examination offers valuable insights into the evolving landscape of religious beliefs and practices in contemporary society.

religion_importance_2014 <- survey_data_2014 %>%
  group_by(qe1) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(qe1), !is.na(qf2)) %>%  # Ensure qe1 and qf2 are not NA
  group_by(qe1, qf2) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(qe1) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(qe1, qf2)

religion_importance_2014_selected <- select(religion_importance_2014, qe1, qf2, Percentage)


religion_importance_2014_wide <- pivot_wider(
  religion_importance_2014_selected,
  names_from = qf2,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(religion_importance_2014_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(religion_importance_2014_wide, "html", align = c('l', rep('c', 10))) %>%
  kable_styling("striped", full_width = F) %>%
  add_header_above(c(" " = 1, "How important is religion in your life (2014) " = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	How important is religion in your life (2014)
Religion	(VOL) Don’t Know/Refused	Not at all important	Not too important	Somewhat important	Very important
Agnostic (not sure if there is a God)	0.8819539	52.1031208	31.0719132	12.483039	3.459973
Atheist (do not believe in God)	0.4553734	82.6047359	10.2003643	4.735883	2.003643
Buddhist	2.2727273	12.1212121	18.1818182	34.090909	33.333333
Christian (VOL)	1.8666667	4.0000000	7.8666667	25.066667	61.200000
Don’t Know/Refused (VOL)	16.8831169	25.3246753	13.6363636	20.129870	24.025974
Hindu	1.5075377	6.0301508	15.5778894	48.743719	28.140704
Jehovah’s Witness (VOL)	0.8196721	0.8196721	0.8196721	6.147541	91.393443
Jewish (Judaism)	0.5903188	9.9173554	19.5985832	36.953955	32.939788
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	0.0000000	0.9036145	2.4096386	12.500000	84.186747
Muslim (Islam)	1.2658228	2.5316456	9.2827004	21.940928	64.978903
Nothing in particular	1.1637239	31.2399679	27.1669342	25.321027	15.108347
Orthodox (Greek, Russian, or some other orthodox church)	0.5376344	3.2258065	11.8279570	33.870968	50.537634
Protestant	0.5870471	1.2435299	4.6521904	21.537685	71.979548
Roman Catholic	0.3891051	2.2234575	7.9071707	30.864369	58.615898
Spiritual but not religious (‘believe in God, but…’) (VOL)	1.5625000	32.0312500	16.4062500	28.906250	21.093750
Unitarian (Universalist) (VOL)	1.8181818	9.6969697	24.2424242	36.363636	27.878788

# Assuming the data frames are named religion_importance_2007 and religion_importance_2014


# Filter both data frames to include only common religions
religion_counts_2007 <- religion_counts_2007 %>%
  filter(Religion %in% common_religions)

religion_counts_2014 <- religion_counts_2014 %>%
  filter(Religion %in% common_religions)

# Convert data from wide to long format for easier plotting
religion_importance_2007_long <- pivot_longer(religion_importance_wide, cols = -Religion, names_to = "Importance", values_to = "Percentage_2007")
religion_importance_2014_long <- pivot_longer(religion_importance_2014_wide, cols = -Religion, names_to = "Importance", values_to = "Percentage_2014")

# Filter both data frames to include only common religions
religion_importance_2007_long <- religion_importance_2007_long %>%
  filter(Religion %in% common_religions)

religion_importance_2014_long <- religion_importance_2014_long %>%
  filter(Religion %in% common_religions)

# Merge the two datasets

religion_importance_combined <- full_join(religion_importance_2007_long, religion_importance_2014_long, by = c("Religion", "Importance"))

# Check and clean any NA values that might result from the join if there are religions not present in both datasets
religion_importance_combined <- drop_na(religion_importance_combined)

# Changes in religion importance 
religion_importance_combined <- religion_importance_combined %>% 
  mutate(Percentage_change = Percentage_2014 - Percentage_2007)


# Heatmap for both years
p <- ggplot(religion_importance_combined, aes(x = Religion, y = Importance, fill = Percentage_change)) +
  geom_tile() + # Create the tiles for the heatmap
  scale_fill_gradient(low = "white", high = "steelblue") + # Gradient from white (low) to steel blue (high)
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1), axis.text.y = element_text(size = 12), title = element_text(size = 20)) +
  labs(title = "Changes in responses on the Importance of Religion by Religious Group",
       x = "Religion",
       y = "Importance",
       fill = "Percentage")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 750)

# Show the interactive plot in your default browser

interactive_plot

Explanation: This visualization illustrates the evolving perception of the importance of religion in the lives of various religious groups between 2007 and 2014. The chart depicts the percentage of respondents within each religious group who reported different levels of importance

Key observations from the visualization might include: 1. Overall, there appears to be a trend towards decreased importance of religion across most religious groups from 2007 to 2014. 2. Some groups, such as atheists and agnostics, show a notable increase in the percentage of respondents who view religion as “Not at all important,” suggesting a shift towards secularism. 3. Conversely, certain groups like Jehovah’s Witnesses and Mormons seem to maintain a consistently high percentage of respondents who perceive religion as “Very important” throughout both years. 4. The chart also highlights variations in the importance of religion among different religious communities, reflecting the diversity of attitudes and beliefs within and across religious traditions.

Shifting Views on Religion and “Culture War” Issues

Shifting Views on Abortion Abortion

This section presents a comparative examination of attitudes towards abortion across different religious groups between the years 2007 and 2014. The data illustrate respondents’ opinions on whether abortion should be illegal or legal in varying circumstances, including “Illegal in all cases,” “Illegal in most cases,” “Legal in all cases,” and “Legal in most cases.” By juxtaposing the percentages of respondents within each religious category, the chart highlights evolving perspectives on abortion rights over time.

Abortion_2014 <- survey_data_2014 %>%
  group_by(qe1) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(qe1), !is.na(qb21)) %>%  # Ensure qe1 and qb21 are not NA
  group_by(qe1, qb21) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(qe1) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(qe1, qb21)

Abortion_2014_selected <- select(Abortion_2014, qe1, qb21, Percentage)


Abortion_2014_wide <- pivot_wider(
  Abortion_2014_selected,
  names_from = qb21,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Abortion_2014_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Abortion_2014_wide, "html", align = c('l', rep('c', 10))) %>%
  kable_styling("striped", full_width = F) %>%
  add_header_above(c(" " = 1, "Abortion Should Be… " = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Abortion Should Be…
Religion	(VOL) Don’t know/Refused	Illegal in all cases	Illegal in most cases	Legal in all cases	Legal in most cases
Agnostic (not sure if there is a God)	2.103121	1.7639077	8.073270	38.941655	49.11805
Atheist (do not believe in God)	2.185792	2.5500911	5.828780	51.639344	37.79599
Buddhist	1.136364	2.6515152	9.848485	40.530303	45.83333
Christian (VOL)	9.333333	17.2000000	23.200000	17.866667	32.40000
Don’t Know/Refused (VOL)	15.584416	5.1948052	11.038961	38.961039	29.22078
Hindu	4.522613	10.5527638	13.065327	26.130653	45.72864
Jehovah’s Witness (VOL)	6.557377	53.6885246	21.721311	4.508197	13.52459
Jewish (Judaism)	2.007084	2.5974026	8.736718	42.739079	43.91972
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	2.861446	9.3373494	62.048193	6.024096	19.72892
Muslim (Islam)	7.594937	12.2362869	24.894515	14.345992	40.92827
Nothing in particular	4.514446	7.8852327	16.512841	31.741573	39.34591
Orthodox (Greek, Russian, or some other orthodox church)	3.763441	8.0645161	33.333333	20.430107	34.40860
Protestant	4.336574	16.9296806	33.385936	13.704078	31.64373
Roman Catholic	4.877710	17.7181768	29.335742	16.133963	31.93441
Spiritual but not religious (‘believe in God, but…’) (VOL)	10.156250	3.9062500	17.187500	30.468750	38.28125
Unitarian (Universalist) (VOL)	3.030303	0.6060606	5.454546	46.060606	44.84848

# Convert data from wide to long format for easier plotting
Abortion_2007_long <- pivot_longer(Abortion_wide, cols = -Religion, names_to = "Importance", values_to = "Percentage_2007")
Abortion_2014_long <- pivot_longer(Abortion_2014_wide, cols = -Religion, names_to = "Importance", values_to = "Percentage_2014")

# Filter both data frames to include only common religions
Abortion_2007_long <- Abortion_2007_long %>%
  filter(Religion %in% common_religions)

Abortion_2014_long <- Abortion_2014_long %>%
  filter(Religion %in% common_religions)
# Merge the two datasets

Abortion_combined <- full_join(Abortion_2007_long, Abortion_2014_long, by = c("Religion","Importance"))

# Check and clean any NA values that might result from the join if there are religions not present in both datasets
Abortion_combined <- drop_na(Abortion_combined)

# Changes in religion importance 
Abortion_combined <- Abortion_combined %>% 
  mutate(Percentage_change = Percentage_2014 - Percentage_2007)


# Heatmap for both years
p <- ggplot(Abortion_combined, aes(x = Religion, y = Importance, fill = Percentage_change)) +
  geom_tile() + # Create the tiles for the heatmap
  scale_fill_gradient(low = "white", high = "steelblue") + # Gradient from white (low) to steel blue (high)
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1), title = element_text(size = 15)) +
  labs(title = "Changes in responses on the Abortion by Religious Group",
       x = "Religion",
       y = "",
       fill = "Percentage")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 750)

# Show the interactive plot in your default browser

interactive_plot

1. Consistency in Views: Some religious groups, such as Mormons (LDS) and Jews, maintained relatively consistent stances on abortion between 2007 and 2014. These groups showed minimal changes in their percentages across the categories of “Illegal in all cases,” “Illegal in most cases,” “Legal in all cases,” and “Legal in most cases.” This consistency suggests a stable adherence to their respective religious teachings or moral convictions regarding abortion over the years.

2. Decrease in Opposition: In several religious categories, there was a decrease in the percentage of respondents favoring strict restrictions on abortion. For example, among Protestants and Roman Catholics, there was a decline in the proportion of respondents advocating for abortion to be illegal in all or most cases. This shift may indicate a growing acceptance of more liberal abortion policies within these religious communities over time.

3. Increase in Support for Legalization: Conversely, some religious groups saw an increase in the percentage of respondents supporting legalized abortion, particularly in certain circumstances. This trend was evident among Atheists, Buddhists, and Muslims, where a higher proportion of respondents expressed support for abortion being legal in all or most cases in 2014 compared to 2007. This suggests a potential loosening of traditional views on abortion within these religious communities.

4. Varied Responses: While some religious groups exhibited clear shifts towards either more restrictive or more permissive attitudes on abortion, others showed mixed or inconsistent changes. For instance, Jehovah’s Witnesses and Orthodox Christians displayed fluctuating percentages across different categories, indicating a divergence of opinions within these religious communities over the years.

Changing Attitudes Towards Homosexuality

This analysis examines the evolving perceptions of homosexuality within different religious communities over the span of seven years, from 2007 to 2014. By comparing data from these two time points, we gain insights into the shifting attitudes towards LGBTQ+ rights and acceptance within various religious groups. The findings highlight trends such as decreasing disapproval, increasing acceptance, and the variability of attitudes across different religious affiliations.

Homosexuality_2014 <- survey_data_2014 %>%
  group_by(qe1) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(qe1), !is.na(qb22)) %>%  # Ensure qe1 and q2b2 are not NA
  group_by(qe1, qb22) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(qe1) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(qe1, qb22)

Homosexuality_2014_selected <- select(Homosexuality_2014, qe1, qb22, Percentage)


Homosexuality_2014_wide <- pivot_wider(
  Homosexuality_2014_selected,
  names_from = qb22,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Homosexuality_2014_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Homosexuality_2014_wide, "html", align = c('l', rep('c', 10))) %>%
  kable_styling("striped", full_width = F) %>%
  add_header_above(c(" " = 1, "Allowing gays and lesbians to marry legally? " = 4, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Allowing gays and lesbians to marry legally?
Religion	(VOL) Don’t know/Refused	Favor	Oppose	Strongly favor	Strongly oppose
Agnostic (not sure if there is a God)	3.188602	32.225237	4.545454	58.208955	1.831750
Atheist (do not believe in God)	2.367942	27.504554	2.914390	65.209472	2.003643
Buddhist	3.030303	39.772727	4.924242	47.727273	4.545454
Christian (VOL)	13.866667	25.466667	19.600000	20.933333	20.133333
Don’t Know/Refused (VOL)	16.883117	26.623377	12.987013	36.363636	7.142857
Hindu	11.055276	33.668342	14.572864	29.648241	11.055276
Jehovah’s Witness (VOL)	10.655738	7.377049	35.245902	5.327869	41.393443
Jewish (Judaism)	4.604486	32.231405	7.674144	49.114522	6.375443
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	6.174699	15.361446	36.295181	7.981928	34.186747
Muslim (Islam)	6.329114	25.738397	22.362869	18.565401	27.004219
Nothing in particular	7.423756	35.433387	11.898074	37.700642	7.544141
Orthodox (Greek, Russian, or some other orthodox church)	6.989247	29.569893	25.806452	26.881720	10.752688
Protestant	7.713673	23.930059	27.269284	14.467870	26.619114
Roman Catholic	8.949416	35.575320	22.206782	20.094497	13.173986
Spiritual but not religious (‘believe in God, but…’) (VOL)	12.500000	32.031250	6.250000	43.750000	5.468750
Unitarian (Universalist) (VOL)	2.424242	24.848485	2.424242	69.090909	1.212121

1. Consistency in Acceptance: In both 2007 and 2014, Agnostics, Atheists, and Buddhists consistently showed high levels of acceptance towards homosexuality. These groups maintained relatively stable percentages of respondents who believed that homosexuality is a way of life that should be accepted.

2. Decrease in Disapproval: Across various religious categories, there was a notable decrease in the percentage of respondents who believed that homosexuality should be discouraged. For example, among Mormons (LDS) and Muslims, there was a decline in the proportion of individuals advocating for discouraging homosexuality. This suggests a shift towards more accepting attitudes within these religious communities over time.

3. Increase in Acceptance: Conversely, some religious groups witnessed an increase in the percentage of respondents supporting the acceptance of homosexuality. Notably, Protestants and Orthodox Christians showed a rise in the proportion of individuals endorsing acceptance of homosexuality as a way of life. This indicates a potential evolution in attitudes towards LGBTQ+ rights within these religious communities over the surveyed period.

4. Religious Variability: The data also reveal a diverse range of attitudes towards homosexuality across different religious groups. While some communities, such as Unitarian Universalists, overwhelmingly support the acceptance of homosexuality, others, like Jehovah’s Witnesses, maintain more conservative views on the matter.

Shifts in Beliefs on Evolution

This comparative study examines the changing attitudes towards evolution across different religious affiliations between 2007 and 2014. The data reveals shifts in opinions regarding the origins of human life and the role of evolution, from the influence of a supreme being to natural processes such as natural selection.

Evolution_2014 <- survey_data_2014 %>%
  group_by(qe1) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(qe1), !is.na(qb30b)) %>%  # Ensure qe1 and q2b2 are not NA
  group_by(qe1, qb30b) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(qe1) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(qe1, qb30b)

Evolution_2014_selected <- select(Evolution_2014, qe1, qb30b, Percentage)


Evolution_2014_wide <- pivot_wider(
  Evolution_2014_selected,
  names_from = qb30b,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Evolution_2014_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Evolution_2014_wide, "html", align = c('l', rep('c', 10))) %>%
  kable_styling("striped", full_width = F) %>%
  add_header_above(c(" " = 1, "Allowing gays and lesbians to marry legally? " = 2, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	Allowing gays and lesbians to marry legally?
Religion	(VOL) Don’t know/Refused	A supreme being guided the evolution of living things for the purpose of creating humans and other life…	Humans and other living things have evolved due to natural processes such as natural selection, OR
Agnostic (not sure if there is a God)	4.812456	6.440198	88.74735
Atheist (do not believe in God)	1.238095	1.904762	96.85714
Buddhist	5.603448	12.068965	82.32759
Christian (VOL)	12.235294	51.058824	36.70588
Don’t Know/Refused (VOL)	20.175439	17.543860	62.28070
Hindu	5.844156	17.532467	76.62338
Jehovah’s Witness (VOL)	0.000000	68.888889	31.11111
Jewish (Judaism)	7.412587	17.622378	74.96504
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	5.797101	69.565217	24.63768
Muslim (Islam)	6.716418	48.507463	44.77612
Nothing in particular	6.381897	21.179761	72.43834
Orthodox (Greek, Russian, or some other orthodox church)	6.034483	43.103448	50.86207
Protestant	7.142857	56.100661	36.75648
Roman Catholic	7.121058	48.809766	44.06918
Spiritual but not religious (‘believe in God, but…’) (VOL)	9.375000	44.791667	45.83333
Unitarian (Universalist) (VOL)	5.298013	9.271523	85.43046

1. Agnostic and Atheist Perspectives: Both agnostic and atheist respondents consistently leaned towards acceptance of evolution due to natural processes across both years. While there were slight fluctuations, the majority maintained their stance in favor of natural selection as the explanation for the origins of human life.

2. Buddhist, Hindu, and Unitarian Universalist Views: These groups also exhibited strong support for evolution driven by natural processes, with minimal changes observed between 2007 and 2014. Their beliefs remained largely consistent over the seven-year period.

3. Christian Denominations: Within Christian denominations, there was a notable divide. While some denominations, such as Unitarian Universalists, showed overwhelming support for evolution driven by natural processes, others, particularly Jehovah’s Witnesses and Mormons, maintained beliefs in a supreme being guiding evolution. Protestant and Roman Catholic groups demonstrated mixed views, with a significant portion supporting natural selection but also a sizable minority endorsing a divine-guided evolution.

4. Overall Trends: Across most religious groups, there was a slight increase in acceptance of evolution driven by natural processes from 2007 to 2014. This indicates a gradual shift towards secular interpretations of evolutionary theory within various religious communities, although some groups still maintain strong beliefs in divine involvement in the process.

Changes in Frequency of Experiencing Spiritual Peace and Well-being

This section examines the shifts in the frequency of experiencing spiritual peace and well-being among various religious groups between 2007 and 2014. The data reveal changes in the proportions of respondents within each group who reported experiencing spiritual peace and well-being at different frequencies, ranging from at least once a week to never.

Peace_Wonder_2014 <- survey_data_2014 %>%
  group_by(qe1) %>% 
  filter(n() > 100) %>% # Keep only religions with more than 100 respondents
  ungroup() %>% 
  filter(!is.na(qe1), !is.na(qi4a)) %>%  # Ensure qe1 and q14a are not NA
  group_by(qe1, qi4a) %>%
  summarise(Count = n(), .groups = 'drop') %>%  # Count occurrences and drop grouping
  ungroup() %>% 
  group_by(qe1) %>%
  mutate(Total = sum(Count), 
         Percentage = (Count / Total) * 100) %>%
  ungroup() %>% 
  arrange(qe1, qi4a)


Peace_Wonder_2014_selected <- select(Peace_Wonder_2014, qe1, qi4a, Percentage)


Peace_Wonder_2014_wide <- pivot_wider(
  Peace_Wonder_2014_selected,
  names_from = qi4a,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)

colnames(Peace_Wonder_2014_wide)[1] <- "Religion"

# Create the table with kable and add styling with kableExtra
table <- kable(Peace_Wonder_2014_wide, "html") %>%
  kable_styling("striped", full_width = T) %>%
  add_header_above(c(" " = 1, "feel a deep sense of spiritual peace and well-being" = 5, " " = 1)) %>%
  row_spec(0, bold = T, color = "white", background = "#636363") %>%
  scroll_box( height = "100%") %>% 
  column_spec(1, bold = T) 

# If you want to see the table in RStudio's Viewer
table

	feel a deep sense of spiritual peace and well-being
Religion	(VOL) Don’t Know/Refused	At least once a week	Never	Once or twice a month	Seldom	Several times a year
Agnostic (not sure if there is a God)	1.4925373	37.51696	13.4328358	18.249661	16.824966	12.483039
Atheist (do not believe in God)	2.2768670	32.69581	28.8706740	14.480874	14.298725	7.377049
Buddhist	0.7575758	62.87879	2.2727273	16.287879	8.712121	9.090909
Christian (VOL)	2.5333333	62.00000	4.1333333	16.800000	7.333333	7.200000
Don’t Know/Refused (VOL)	13.6363636	46.75325	9.7402597	11.038961	10.389610	8.441558
Hindu	2.0100503	45.22613	2.0100503	16.582915	13.567839	20.603015
Jehovah’s Witness (VOL)	3.2786885	82.78689	1.2295082	4.918033	2.868853	4.918033
Jewish (Judaism)	1.7709563	37.78040	9.2089728	17.709563	18.536009	14.994097
Mormon (Church of Jesus Christ of Latter-day Saints/LDS)	1.0542169	82.07831	1.3554217	8.734940	2.560241	4.216868
Muslim (Islam)	3.7974684	64.13502	0.8438819	15.189873	5.485232	10.548523
Nothing in particular	2.1268058	42.85714	11.1757624	15.308989	17.475923	11.055377
Orthodox (Greek, Russian, or some other orthodox church)	1.0752688	55.37634	4.3010753	20.967742	8.602151	9.677419
Protestant	1.1551572	69.63136	1.8621386	13.142280	6.590077	7.618988
Roman Catholic	1.3618677	57.99055	3.1267371	16.286826	10.241801	10.992218
Spiritual but not religious (‘believe in God, but…’) (VOL)	0.7812500	75.00000	1.5625000	12.500000	5.468750	4.687500
Unitarian (Universalist) (VOL)	3.0303030	56.36364	3.0303030	13.939394	12.727273	10.909091

# Convert data from wide to long format for easier plotting
Peace_Wonder_2007_long <- pivot_longer(Peace_Wonder_wide, cols = -Religion, names_to = "Importance", values_to = "Percentage_2007")
Peace_Wonder_2014_long <- pivot_longer(Peace_Wonder_2014_wide, cols = -Religion, names_to = "Importance", values_to = "Percentage_2014")

# Filter both data frames to include only common religions
Peace_Wonder_2007_long <- Peace_Wonder_2007_long %>%
  filter(Religion %in% common_religions)

Peace_Wonder_2014_long <- Peace_Wonder_2014_long %>%
  filter(Religion %in% common_religions)
# Merge the two datasets

Peace_Wonder_combined <- full_join(Peace_Wonder_2007_long, Peace_Wonder_2014_long, by = c("Religion","Importance"))

# Check and clean any NA values that might result from the join if there are religions not present in both datasets
Peace_Wonder_combined <- drop_na(Peace_Wonder_combined)

# Changes in religion importance 
Peace_Wonder_combined <- Peace_Wonder_combined %>% 
  mutate(Percentage_change = Percentage_2014 - Percentage_2007)


# Heatmap for both years
p <- ggplot(Peace_Wonder_combined, aes(x = Religion, y = Importance, fill = Percentage_change)) +
  geom_tile() + # Create the tiles for the heatmap
  scale_fill_gradient(low = "white", high = "steelblue") + # Gradient from white (low) to steel blue (high)
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1), title = element_text(size = 15)) +
  labs(title = "Changes in responses on the sense spritual by Religious Group",
       x = "Religion",
       y = "",
       fill = "Percentage")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 750)

# Show the interactive plot in your default browser

interactive_plot

1. Overall Trends: Across most religious groups, there is a consistent pattern of a decrease in the proportion of respondents who report experiencing spiritual peace and well-being at least once a week, coupled with an increase in those reporting never experiencing it. This suggests a potential decline in regular spiritual practices or feelings of spiritual fulfillment over the surveyed period.

2. Variability Among Groups: While the overall trend shows a decrease in weekly spiritual experiences, the extent of this change varies among different religious groups. For instance, some groups, such as Jehovah’s Witnesses and Mormons, maintained high levels of weekly spiritual experiences, while others, like Buddhists and Hindus, showed more variability.

3. Shifts in “Never” Category: A significant increase in the proportion of respondents reporting never experiencing spiritual peace and well-being is notable across several religious groups. This shift raises questions about potential changes in religious participation, beliefs, or societal factors impacting spiritual practices.

4. Stability in Some Groups: Interestingly, certain groups, such as Muslims and Orthodox Christians, show relatively stable patterns in the frequency of spiritual experiences. This stability may reflect the resilience of traditional religious practices and beliefs within these communities over time.

Conclusion

Throughout Chapter 3, we have systematically examined the shifting landscapes of religious affiliation, the importance of religion, attitudes towards controversial cultural issues, and experiences of spiritual well-being. Our analyses span from broad demographic shifts to nuanced changes in beliefs and practices over time.

Changes in Religious Affiliations: - The analysis reveals a significant fluidity in religious affiliations between 2007 and 2014. Notably, there was a discernible trend towards non-affiliation, with increases in the “Nothing in particular” category, alongside rises in atheism and agnosticism. Simultaneously, traditional religious affiliations such as Protestantism and Catholicism saw declines, reflecting a broader trend of secularization in American society.

Changing Perceptions of Religion’s Importance: - The importance of religion has seen varying levels of decline across different groups, indicating a shift towards secular values over the surveyed period. However, contrasts remain stark among different religious communities, with some maintaining a strong emphasis on religious practices and beliefs.

Shifting Views on Religion and ‘Culture War’ Issues: - Attitudes towards contentious social issues like abortion and homosexuality have evolved, with a general trend towards more liberal views, especially among younger and less traditionally religious demographics. However, there remains substantial variability, with some groups showing remarkable consistency in their views due to deep-seated religious beliefs.

Changes in Frequency of Experiencing Spiritual Peace and Well-being: - There appears to be a general decline in regular spiritual experiences, with an increase in individuals reporting never experiencing spiritual peace and well-being. This change could reflect shifting engagement with traditional religious practices or broader societal changes affecting personal and communal spiritual lives.

Chapter 4 : Modelings

Modeling Religious Affiliation Dynamics: A Transition Probability Approach

For Chapter 4 of your project, you are conducting an exploration of religious dynamics, focusing specifically on how individuals transition between religions from childhood to adulthood. Your analysis aims to model the changing composition of religious affiliations in a society over time, using a matrix of transition probabilities. Here’s a structured and clear explanation of your approach and objectives:

Purpose of the Analysis

The primary goal of this analysis is to understand the dynamics of religious affiliation in society by studying how individuals may change their religious affiliations from childhood to adulthood. You seek to:

1. Quantify the Transition Probabilities: Calculate the probability that a person born or identified with one religion (childhood religion) will transition to another religion (adulthood religion). This involves analyzing survey data to determine the proportion of individuals who switch from each religion to another, compared to the total number of individuals who started in each religion.

2. Simulate Religious Composition Over Time: Use the calculated transition probabilities to simulate how the distribution of religious affiliations changes over time within a fixed population. This simulation will help visualize potential trends and shifts in religious composition under the assumption that no other demographic factors (like birth, death, or immigration) affect the population.

3. Identify Steady States: Determine if the simulation reaches a steady state where the proportions of each religion in the population remain constant over successive time periods, suggesting a balance in how individuals transition between religions.

Methodology

Calculation of Transition Probabilities

For each religion X (childhood religion), calculate the probability of transitioning to each religion Y (adulthood religion) as follows:

\[ P(Y|X) = \frac{\text{Number of people who changed their religion from } X \text{ to } Y}{\text{Total number of children with religion } X} \]

This results in a transition matrix where each element (i, j) represents the probability of transitioning from religion i to religion j.

Simulation of Religious Composition

• Initial Setup: Start with the initial distribution of religious affiliations based on the survey data.
• Application of Transition Matrix: At each simulation step, apply the transition matrix to the current distribution of religions to obtain the distribution for the next time period.
• Repeat Until Steady State: Continue applying the transition matrix over multiple time periods to observe how the religious composition evolves until it stabilizes or reaches a steady state.

# the columns are named 'qj1' for childhood religion and 'qe1' for current religion
transition_counts <- survey_data_2014 %>%
  filter(!is.na(qj1) & !is.na(qe1)) %>% 
  group_by(qj1) %>% 
  filter(n()>100) %>% 
  ungroup() %>% 
  group_by(qe1) %>% 
  filter(n()>150) %>% 
  ungroup() %>%
  group_by(qj1,qe1) %>%
  filter(n()>10) %>%
  summarise(count = n()) %>%
  ungroup()

# Summarize total counts for each childhood religion
total_per_child_religion <- transition_counts %>%
  group_by(qj1) %>%
  summarise(Total_Count = sum(count))

# Calculate percentages
transition_percentages <- transition_counts %>%
  left_join(total_per_child_religion, by = "qj1") %>%
  mutate(Percentage = (count / Total_Count) * 100) %>%
  select(qj1, qe1, count, Percentage)


# Create nodes and links for the Sankey diagram
node_names <- unique(c(as.character(transition_counts$qj1), as.character(transition_counts$qk1)))
nodes <- data.frame(name = node_names)
links <- transition_counts %>%
  mutate(source = match(qj1, node_names) - 1,  # Indexing starts at 0 for JavaScript
         target = match(qe1, node_names) - 1,
         value = count)


# First we need to create an igraph object from the transition data
graph_data <- transition_percentages %>%
  select(source = qj1, target = qe1, weight = Percentage)



# Create the graph object using igraph
graph <- graph_from_data_frame(transition_percentages, directed = TRUE)

# Create groups based on connected components if there is no explicit community detection
com <- components(graph)
grp <- com$membership

# Add group information to the nodes/vertices
V(graph)$group <- grp



mycolors <- RColorBrewer::brewer.pal(min(max(grp), 8), "Set1")
if (max(grp) > 8) {
  mycolors <- grDevices::colors()[seq(from = 1, to = max(grp))]
}
names(mycolors) <- unique(grp)

# Make the arc diagram with ggraph
p <- ggraph(graph, layout = 'linear') + 
  geom_edge_arc(edge_width = 0.3, edge_colour="black", edge_alpha=0.2, fold=TRUE) +
  geom_node_point(aes(size = degree(graph, v = V(graph)), color = factor(group)), alpha = 0.5) +
  scale_size_continuous(range=c(0.5,8)) +
  scale_color_manual(values=mycolors) +
  geom_node_text(aes(label=name), angle=65, hjust=1, nudge_y = -1.1, size=2.3) +
  labs(title = 'transition of religions')+
  theme_void() +
  theme(
    title = element_text(size = 20),
    legend.position="none",
    plot.margin=unit(c(0,0,0.4,0), "null"),
    panel.spacing=unit(c(0,0,3.4,0), "null")
  ) +
  expand_limits(x = c(-1.2, 1.2), y = c(-5.6, 1.2)) 

# Use ggsave() to save the plot with specified dimensions
ggsave("arc_diagram.png", plot = p, width = 12, height = 8, units = "in", dpi = 300)

This arc diagram visually represents the transitions individuals make from their childhood religion to their current religion. Each node (dot) on the horizontal axis represents a different religious affiliation, and the arcs (curved lines) indicate the movement of people between these affiliations over time. The size of a node reflects the number of individuals who either remain in or leave a particular religion.

transition_percentages_2014_selected <- select(transition_percentages, qj1, qe1, Percentage)


transition_percentages_2014_wide <- pivot_wider(
  transition_percentages_2014_selected,
  names_from = qj1,  # This column's unique values will become the names of the new columns
  values_from = Percentage,  # The cell values will come from the 'Percentage' column
  values_fill = list(Percentage = 0)  # Fill missing combinations with 0%
)
transition_percentages_2014_wide

# Heatmap for both years
p <- ggplot(transition_percentages_2014_selected, aes(x = qe1, y = qj1, fill = Percentage)) +
  geom_tile() + # Create the tiles for the heatmap
  scale_fill_gradient(low = "white", high = "steelblue") + # Gradient from white (low) to steel blue (high)
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1), title = element_text(size = 15)) +
  labs(title = "Transition Probability of changing religion",
       x = "Adulthood Religion",
       y = "Childhood Religion",
       fill = "Percentage")

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 750)

# Show the interactive plot in your default browser

interactive_plot

The heatmap depicts changes in religious affiliation from childhood to adulthood across various groups. Observing the concentration of darker shades along the diagonal would suggest that many individuals remain within their childhood religion. However, deviations from the diagonal where darker shades appear indicate notable transitions from one religion to another (notable for Protestant and Nothing in particular).

To simulate the evolution of the population’s religious affiliations, we start with an initial population vector \(\mathbf{P}(0)\), where each element \(P_i(0)\) represents the number of individuals affiliated with religion \(i\) at the beginning of our simulation (time \(t = 0\)).

At each subsequent time point \(t\), we update the population vector using the following formula:

\[ \mathbf{P}(t) = \mathbf{P}(t-1) \times \mathbf{T} \]

where: - \(\mathbf{P}(t)\) is the population vector at time \(t\), with \(P_i(t)\) being the number of individuals affiliated with religion \(i\). - \(\mathbf{P}(t-1)\) is the population vector from the previous time point \(t-1\). - \(\mathbf{T}\) is the transition matrix, with each element \(T_{ij}\) representing the probability of an individual transitioning from religion \(i\) to religion \(j\).

The multiplication of \(\mathbf{P}(t-1)\) by \(\mathbf{T}\) applies the transition probabilities to the population vector, producing a new distribution of religious affiliations at time \(t\). This process models how individuals may change their religious affiliations over time, allowing us to forecast the potential shifts in the religious composition of the population.

By repeating this calculation for each time step, we construct a time series of population vectors \(\{ \mathbf{P}(1), \mathbf{P}(2), \ldots, \mathbf{P}(t) \}\), which depict the dynamics of religious affiliation within the population across the simulated time periods.

# Convert percentage to probability
transition_matrix <- transition_percentages_2014_wide %>%
  select(-qe1) %>% # Remove the first column which contains labels
  mutate_all(~ .x / 100) # Convert percentages to probabilities


# Convert percentages to probabilities (divide by 100)
transition_percentages_2014_selected$Percentage <- transition_percentages_2014_selected$Percentage / 100

# Creating a wide matrix with all religions both in rows and columns
transition_matrix <- transition_percentages_2014_selected %>%
  spread(key = qj1, value = Percentage, fill = 0)  # Use the correct key and value fields

# Ensure all unique religions are included as rows and columns
all_religions <- union(unique(transition_percentages_2014_selected$qe1), unique(transition_percentages_2014_selected$qj1))

# Create an empty full transition matrix with zeros
full_matrix <- matrix(0, nrow = length(all_religions), ncol = length(all_religions),
                      dimnames = list(all_religions, all_religions))

# Fill the full matrix with the available probabilities
for (i in 1:nrow(transition_percentages_2014_selected)) {
  from_religion <- transition_percentages_2014_selected$qe1[i]
  to_religion <- transition_percentages_2014_selected$qj1[i]
  probability <- transition_percentages_2014_selected$Percentage[i]
  full_matrix[from_religion, to_religion] <- probability
}

# Convert to a data frame for visualization/checking
transition_df <- as.data.frame(full_matrix)


simulate_religion_dynamics <- function(pop, P, max_time) {
  pop_matrix <- matrix(0, nrow = max_time, ncol = length(pop))
  pop_matrix[1,] <- pop
  for (t in 2:max_time) {
    pop_matrix[t,] <- pop_matrix[t-1,] %*% P
  }
  return(pop_matrix)
}

# Initial Population based on the data
initial_population <- c(1474 , 1098 , 5000, 15000, 264, 750, 7200, 200, 199, 244, 847, 664, 237, 186)

# Run the simulation
result <- simulate_religion_dynamics(initial_population, t(full_matrix), max_time = 40)

# Assuming 'result' is your matrix from the simulation
result_df <- as.data.frame(result)
time <- 1:nrow(result_df)
result_df <- cbind(time, result_df)

# Melt the data frame for ggplot
result_long <- result_df %>%
  pivot_longer(cols = -time, names_to = "Religion", values_to = "Population")

# # Name the religions appropriately if not automatically named
religions <- names(transition_df)
result_long$Religion <- rep(religions, times = 40)


p <- ggplot(result_long, aes(x = time, y = Population, color = Religion )) +
  geom_line() +
  theme_minimal() +
  labs(title = "Religion Population Dynamics",
       x = "Time",
       y = "Population") +
  
  theme(legend.title = element_blank())


# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 600)

# Show the interactive plot in your default browser

interactive_plot

Conclusion

The graph shows us a snapshot of religious shifts over time. We observe that the number of individuals identifying as Protestant and those stating “Nothing in particular” are on the rise, while the Roman Catholic group is seeing a decrease. Other religions appear to be fairly steady in their numbers. These trends are mostly consistent with the actual changes in religious affiliation observed between 2007 and 2014, with the growth of the Protestant group as an exception.

It’s important to note that this is a simplified model that doesn’t account for factors like immigration, birth, and mortality rates, which all play a significant role in shaping the religious composition of a population. Nonetheless, this model serves as a promising starting point for more detailed future studies that could incorporate these additional elements.

Analyzing the Predictors of Religious Change from Childhood to Adulthood Using Random Forest and GLM

Random Forest Classifier

Understanding how and why individuals change their religious beliefs over time is crucial for both sociological research and practical applications within community planning and policy making. The complexity of religious belief systems and their influence on personal and social life demand a sophisticated analytical approach to capture the nuanced changes individuals may experience.

# Calculate the percentage of missing values for each column
missing_percentage <- sapply(survey_data_2014, function(x) sum(is.na(x)) / length(x)) * 100

# Set threshold for removal
threshold <- 0

# Find columns with missing value percentage greater than the threshold
columns_to_remove <- names(missing_percentage[missing_percentage > threshold])

# Remove the columns from the dataset
data_clean_2014 <- survey_data_2014[, !(names(survey_data_2014) %in% columns_to_remove)]

data_clean_2014 <- data_clean_2014 %>%
  mutate(Religion_Change = if_else(qj1 == qe1, "Same", "Changed")) %>% 
  filter(respondent_birthregion != "Don't know/refused") %>%  # Remove "Don't know/refused"
  mutate(Immigration_Status = if_else(respondent_birthregion == "U.S.", "Non-Immigrant", "Immigrant"))

# # Remove feature
data_clean_2014 <- select(data_clean_2014,-qj1, -qe1)

# # Ensure the target variable is a factor
data_clean_2014$Religion_Change <- factor(data_clean_2014$Religion_Change)

data_clean_2014 <- data_clean_2014 %>% 
  select( -RELTRAD, -CHRELTRAD, -CHFAMILY, -CHDENOM, -CHPROTFAM, -FAMILY, -PROTFAM, -DENOM, -WEIGHT, -resp,-...1,)


set.seed(123)  # for reproducibility
indices <- sample(1:nrow(data_clean_2014), size = 0.7 * nrow(data_clean_2014))
train_data <- data_clean_2014[indices, ]
test_data <- data_clean_2014[-indices, ]



rf_classifier <- randomForest(Religion_Change ~ ., data = train_data, ntree = 200)

# Making predictions
predictions <- predict(rf_classifier, test_data)

# Evaluate model performance
conf_matrix <- table(Predicted = predictions, Actual = test_data$Religion_Change)
accuracy <- sum(diag(conf_matrix)) / sum(conf_matrix)
print(paste("Accuracy: ", accuracy))

## [1] "Accuracy:  0.750549345562243"

conf_matrix

##          Actual
## Predicted Changed Same
##   Changed    1683  597
##   Same       2014 6173

#rain the Random Forest classifier using the training data. Assume that all other columns except your target variable (Religion_Change) are predictors.

The accuracy of your Random Forest model is approximately 75.05%, indicating that about three-quarters of the predictions made by the model are correct. The confusion matrix provides a more detailed breakdown of the model’s performance:

• True Positives (TP): 1683 instances where the model correctly predicted ‘Changed’.
• True Negatives (TN): 6173 instances where the model correctly predicted ‘Same’.
• False Positives (FP): 597 instances where the model incorrectly predicted ‘Changed’ when the actual class was ‘Same’.
• False Negatives (FN): 2014 instances where the model incorrectly predicted ‘Same’ when the actual class was ‘Changed’.

# Assuming 'rf' is your trained random forest model
importance_data <- importance(rf_classifier)
importance_df <- as.data.frame(importance_data)
importance_df$Feature <- rownames(importance_df)

# Using ggplot2 to create the importance plot with gradient color
p <- ggplot(importance_df, aes(x = reorder(Feature, MeanDecreaseGini), y = MeanDecreaseGini, fill = MeanDecreaseGini)) +
  geom_col() +  # Removing the specific fill color from geom_col
  scale_fill_gradient(low = "lightblue", high = "dodgerblue") +  # Define gradient from light blue to dodger blue
  coord_flip() +  # Flip coordinates for horizontal layout
  labs(title = "Feature Importance", x = "Features", y = "Increase in Node Purity") +
  theme_minimal()

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 1000)

# Show the interactive plot in your default browser

interactive_plot

In my analysis, numerous features significantly impact the classification results. However, for the next phase of the study, I aim to focus on features that offer clearer interpretability. Specifically, we plan to hone in on variables like income, education, age, political affiliation, marital status, self-reported happiness, and geographic location. These factors are not only straightforward to understand but also tend to have direct and tangible effects on individuals’ life decisions and perspectives.

By narrowing our analysis to these interpretable factors, I hope to see how the results differ when we limit the model to variables that are more conceptually accessible and traditionally relevant in sociological research. Additionally, this approach will allow us to identify which of these easily understandable factors are most influential in predicting changes in religious affiliation, thereby enhancing the practical applicability and relevance of our findings.

data_clean_2014 <- data_clean_2014 %>% 
  select(Religion_Change , agerec, income, educ, state, cregion, marital, qa1, qa2a, qa2b, qf2, attend)

# Renaming columns
data_clean_2014 <- data_clean_2014 %>%
  rename(
    
    religious_importance = qf2,
    happiness = qa1,
    family_life_satisfaction = qa2a,
    health_life_satisfaction = qa2b,

  )

set.seed(123)  # for reproducibility
indices <- sample(1:nrow(data_clean_2014), size = 0.7 * nrow(data_clean_2014))
train_data <- data_clean_2014[indices, ]
test_data <- data_clean_2014[-indices, ]

rf_classifier <- randomForest(Religion_Change ~ ., data = train_data, ntree = 200)

# Making predictions
predictions <- predict(rf_classifier, test_data)

# Evaluate model performance
conf_matrix <- table(Predicted = predictions, Actual = test_data$Religion_Change)
accuracy <- sum(diag(conf_matrix)) / sum(conf_matrix)
print(paste("Accuracy: ", accuracy))

## [1] "Accuracy:  0.717206458393045"

conf_matrix

##          Actual
## Predicted Changed Same
##   Changed    1581  844
##   Same       2116 5926

Even after reducing the dimensionality of our feature set to focus on 11 easily interpretable factors, the results remain quite comparable to the previous broader model. The accuracy of the current model stands at approximately 71.72%, only slightly lower than the earlier model’s accuracy of 75.05%. This suggests that these selected interpretable factors—such as income, education, age, and others—still capture a significant portion of the variability needed to predict changes in religious affiliation effectively. This reduction in complexity, while maintaining a similar level of predictive power, underscores the relevance and robustness of these key factors in our analysis.

# Assuming 'rf' is your trained random forest model
importance_data <- importance(rf_classifier)
importance_df <- as.data.frame(importance_data)
importance_df$Feature <- rownames(importance_df)

# Using ggplot2 to create the importance plot with gradient color
p <- ggplot(importance_df, aes(x = reorder(Feature, MeanDecreaseGini), y = MeanDecreaseGini, fill = MeanDecreaseGini)) +
  geom_col() +  # Removing the specific fill color from geom_col
  scale_fill_gradient(low = "lightblue", high = "dodgerblue") +  # Define gradient from light blue to dodger blue
  coord_flip() +  # Flip coordinates for horizontal layout
  labs(title = "Feature Importance", x = "Features", y = "Increase in Node Purity") +
  theme_minimal()

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 600)

# Show the interactive plot in your default browser

interactive_plot

1. religious_importance (1565.3918): This is the most significant factor, indicating the personal significance of religion to an individual. A higher or lower value here directly impacts one’s likelihood to maintain or change their religious affiliation, reflecting the core of how individuals perceive and prioritize their faith.

2. state (1495.3427): The state where an individual resides is a crucial factor, likely due to the diverse religious landscapes across different states. This can influence exposure to various religious practices and communities, affecting personal religious changes.

3. agerec (1269.5415): Age often correlates with changes in life perspective and priorities. As people age, their life experiences can lead to shifts in religious beliefs and practices, making this a key predictor of religious change.

4. income (1213.9837): Financial status influences many aspects of a person’s life, including the ability to participate in certain religious activities, exposure to diverse cultures, and general life satisfaction, all of which can affect religious affiliation.

5. educ (979.5771): Educational attainment exposes individuals to various worldviews and critical thinking skills, which can lead to questioning and potentially changing religious beliefs.

6. attend (876.4653): The frequency with which individuals attend religious services is both an indicator of current religious engagement and a predictor of future religious adherence or change.

7. health_life_satisfaction (633.0054): How satisfied individuals are with their health can impact their spiritual beliefs, particularly as people often turn towards or away from spirituality in times of health crises.

8. marital (646.6567): Changes in marital status can profoundly affect personal life, including religious practices and beliefs, as marriage can introduce new dynamics and influences regarding religion.

9. happiness (552.6725): Overall happiness might influence how individuals feel about their current religious standings. Satisfied individuals might see no reason for change, whereas those less happy might seek new spiritual paths.

10. cregion (469.2322): Similar to state, the broader cultural and regional setting can influence religious practices and changes, reflecting the social and cultural diversity impacting religious identity.

11. family_life_satisfaction (435.1318): Satisfaction with family life can significantly influence religious stability or change. Family dynamics often play a crucial role in religious upbringing and continued practice.

These factors, listed by their importance, provide a comprehensive view of the determinants most predictive of changes in religious affiliation. Understanding these can help in developing targeted interventions, policies, and further research into the dynamics of religious identity.

GLM Model

In exploring factors that influence changes in religious beliefs, it’s crucial to employ a statistical method that not only effectively handles binary outcomes but also provides interpretable results. This is where Generalized Linear Models (GLM) come into play, offering a robust framework for our analysis.

Why GLM? GLMs are particularly adept at modeling binary outcomes—such as whether an individual has changed their religious affiliation (Yes/No). This capability stems from the model’s use of a logistic function, which naturally maps any input through predictors to a probability between 0 and 1, perfectly suiting binary classification tasks.

Additionally, GLMs bring a high level of interpretability to the table. Each predictor’s effect is quantified with a coefficient that directly corresponds to the change in the log odds of the outcome per unit change in the predictor, holding all other predictors constant. This characteristic makes it easier to understand which factors are most influential in determining religious changes and allows these effects to be directly communicated in terms of odds ratios—a format that is intuitively understandable for both technical and non-technical audiences.

# First make sure to use as.factor for catagorical predictors
data_clean_2014$religious_importance <- as.factor(data_clean_2014$religious_importance)
data_clean_2014$educ <- as.factor(data_clean_2014$educ)
data_clean_2014$income <- as.factor(data_clean_2014$income)  
data_clean_2014$marital <- as.factor(data_clean_2014$marital)
data_clean_2014$attend <- as.factor(data_clean_2014$attend)
data_clean_2014$agerec <- as.factor(data_clean_2014$agerec)



glm_model <- glm(Religion_Change ~ income + educ + agerec + marital + attend + religious_importance   , family = binomial(link = "logit"), data = data_clean_2014)

# Display the summary of the model
summary(glm_model)

## 
## Call:
## glm(formula = Religion_Change ~ income + educ + agerec + marital + 
##     attend + religious_importance, family = binomial(link = "logit"), 
##     data = data_clean_2014)
## 
## Coefficients:
##                                                                                                                          Estimate
## (Intercept)                                                                                                              -1.07512
## income$150,000 or more                                                                                                    0.02589
## income10 to under $20,000                                                                                                -0.14114
## income100 to under $150,000                                                                                               0.03947
## income20 to under $30,000                                                                                                -0.14344
## income30 to under $40,000                                                                                                -0.05627
## income40 to under $50,000                                                                                                -0.01191
## income50 to under $75,000                                                                                                -0.02349
## income75 to under $100,000                                                                                               -0.02663
## incomeLess than $10,000                                                                                                  -0.23246
## educFour year college or university degree/Bachelor's degree (e.g., BS, BA, AB)                                           0.30998
## educHigh school graduate (Grade 12 with diploma or GED certificate)                                                       0.41271
## educHigh school incomplete (Grades 9-11 or Grade 12 with NO diploma)                                                      0.34843
## educLess than high school (Grades 1-8 or no formal schooling)                                                             0.50747
## educPostgraduate or professional degree, including master's, doctorate, medical or law degree (e.g., MA, MS, PhD, MD...)  0.30821
## educSome college, no degree (includes some community college)                                                             0.21628
## educSome postgraduate or professional schooling, no postgraduate degree (e.g. some graduate school)                       0.36987
## educTwo year associate degree from a college or university                                                                0.27343
## agerec35-39                                                                                                              -0.07722
## agerec40-44                                                                                                               0.03336
## agerec45-49                                                                                                               0.16720
## agerec50-54                                                                                                               0.17924
## agerec55-59                                                                                                               0.22840
## agerec60-64                                                                                                               0.30649
## agerec65-69                                                                                                               0.45991
## agerec70-74                                                                                                               0.38036
## agerec75-79                                                                                                               0.54391
## agerec80-84                                                                                                               0.57401
## agerec85-89                                                                                                               0.88778
## agerecAge 24 or younger                                                                                                  -0.03947
## agerecAge 25-29                                                                                                          -0.04880
## agerecAge 90 or older                                                                                                     0.96888
## agerecDon't know/refused                                                                                                  0.27943
## maritalDivorced                                                                                                           0.26950
## maritalLiving with a partner                                                                                              0.17214
## maritalMarried                                                                                                            0.36426
## maritalNever been married                                                                                                 0.45845
## maritalSeparated                                                                                                          0.30206
## maritalWidowed                                                                                                            0.44564
## attendA few times a year                                                                                                  0.22583
## attendMore than once a week                                                                                               0.22907
## attendNever                                                                                                              -0.60065
## attendOnce a week                                                                                                         0.38835
## attendOnce or twice a month                                                                                               0.40638
## attendSeldom                                                                                                             -0.24097
## religious_importanceNot at all important                                                                                 -0.73885
## religious_importanceNot too important                                                                                     0.18073
## religious_importanceSomewhat important                                                                                    0.91826
## religious_importanceVery important                                                                                        1.16394
##                                                                                                                          Std. Error
## (Intercept)                                                                                                                 0.29580
## income$150,000 or more                                                                                                      0.05738
## income10 to under $20,000                                                                                                   0.05555
## income100 to under $150,000                                                                                                 0.05402
## income20 to under $30,000                                                                                                   0.05443
## income30 to under $40,000                                                                                                   0.05560
## income40 to under $50,000                                                                                                   0.05719
## income50 to under $75,000                                                                                                   0.05018
## income75 to under $100,000                                                                                                  0.05251
## incomeLess than $10,000                                                                                                     0.06093
## educFour year college or university degree/Bachelor's degree (e.g., BS, BA, AB)                                             0.19063
## educHigh school graduate (Grade 12 with diploma or GED certificate)                                                         0.19068
## educHigh school incomplete (Grades 9-11 or Grade 12 with NO diploma)                                                        0.19854
## educLess than high school (Grades 1-8 or no formal schooling)                                                               0.20471
## educPostgraduate or professional degree, including master's, doctorate, medical or law degree (e.g., MA, MS, PhD, MD...)    0.19147
## educSome college, no degree (includes some community college)                                                               0.19114
## educSome postgraduate or professional schooling, no postgraduate degree (e.g. some graduate school)                         0.20588
## educTwo year associate degree from a college or university                                                                  0.19274
## agerec35-39                                                                                                                 0.06767
## agerec40-44                                                                                                                 0.06658
## agerec45-49                                                                                                                 0.06546
## agerec50-54                                                                                                                 0.06205
## agerec55-59                                                                                                                 0.06142
## agerec60-64                                                                                                                 0.06199
## agerec65-69                                                                                                                 0.06418
## agerec70-74                                                                                                                 0.06882
## agerec75-79                                                                                                                 0.08112
## agerec80-84                                                                                                                 0.09016
## agerec85-89                                                                                                                 0.12265
## agerecAge 24 or younger                                                                                                     0.06495
## agerecAge 25-29                                                                                                             0.06658
## agerecAge 90 or older                                                                                                       0.17978
## agerecDon't know/refused                                                                                                    0.10781
## maritalDivorced                                                                                                             0.17970
## maritalLiving with a partner                                                                                                0.18310
## maritalMarried                                                                                                              0.17697
## maritalNever been married                                                                                                   0.17926
## maritalSeparated                                                                                                            0.19354
## maritalWidowed                                                                                                              0.18237
## attendA few times a year                                                                                                    0.16225
## attendMore than once a week                                                                                                 0.16324
## attendNever                                                                                                                 0.16426
## attendOnce a week                                                                                                           0.16205
## attendOnce or twice a month                                                                                                 0.16341
## attendSeldom                                                                                                                0.16230
## religious_importanceNot at all important                                                                                    0.13593
## religious_importanceNot too important                                                                                       0.13333
## religious_importanceSomewhat important                                                                                      0.13091
## religious_importanceVery important                                                                                          0.12995
##                                                                                                                          z value
## (Intercept)                                                                                                               -3.635
## income$150,000 or more                                                                                                     0.451
## income10 to under $20,000                                                                                                 -2.541
## income100 to under $150,000                                                                                                0.731
## income20 to under $30,000                                                                                                 -2.635
## income30 to under $40,000                                                                                                 -1.012
## income40 to under $50,000                                                                                                 -0.208
## income50 to under $75,000                                                                                                 -0.468
## income75 to under $100,000                                                                                                -0.507
## incomeLess than $10,000                                                                                                   -3.815
## educFour year college or university degree/Bachelor's degree (e.g., BS, BA, AB)                                            1.626
## educHigh school graduate (Grade 12 with diploma or GED certificate)                                                        2.164
## educHigh school incomplete (Grades 9-11 or Grade 12 with NO diploma)                                                       1.755
## educLess than high school (Grades 1-8 or no formal schooling)                                                              2.479
## educPostgraduate or professional degree, including master's, doctorate, medical or law degree (e.g., MA, MS, PhD, MD...)   1.610
## educSome college, no degree (includes some community college)                                                              1.132
## educSome postgraduate or professional schooling, no postgraduate degree (e.g. some graduate school)                        1.797
## educTwo year associate degree from a college or university                                                                 1.419
## agerec35-39                                                                                                               -1.141
## agerec40-44                                                                                                                0.501
## agerec45-49                                                                                                                2.554
## agerec50-54                                                                                                                2.889
## agerec55-59                                                                                                                3.719
## agerec60-64                                                                                                                4.944
## agerec65-69                                                                                                                7.166
## agerec70-74                                                                                                                5.527
## agerec75-79                                                                                                                6.705
## agerec80-84                                                                                                                6.367
## agerec85-89                                                                                                                7.238
## agerecAge 24 or younger                                                                                                   -0.608
## agerecAge 25-29                                                                                                           -0.733
## agerecAge 90 or older                                                                                                      5.389
## agerecDon't know/refused                                                                                                   2.592
## maritalDivorced                                                                                                            1.500
## maritalLiving with a partner                                                                                               0.940
## maritalMarried                                                                                                             2.058
## maritalNever been married                                                                                                  2.557
## maritalSeparated                                                                                                           1.561
## maritalWidowed                                                                                                             2.444
## attendA few times a year                                                                                                   1.392
## attendMore than once a week                                                                                                1.403
## attendNever                                                                                                               -3.657
## attendOnce a week                                                                                                          2.397
## attendOnce or twice a month                                                                                                2.487
## attendSeldom                                                                                                              -1.485
## religious_importanceNot at all important                                                                                  -5.435
## religious_importanceNot too important                                                                                      1.356
## religious_importanceSomewhat important                                                                                     7.014
## religious_importanceVery important                                                                                         8.957
##                                                                                                                          Pr(>|z|)
## (Intercept)                                                                                                              0.000278
## income$150,000 or more                                                                                                   0.651860
## income10 to under $20,000                                                                                                0.011069
## income100 to under $150,000                                                                                              0.465053
## income20 to under $30,000                                                                                                0.008405
## income30 to under $40,000                                                                                                0.311455
## income40 to under $50,000                                                                                                0.834978
## income50 to under $75,000                                                                                                0.639730
## income75 to under $100,000                                                                                               0.612022
## incomeLess than $10,000                                                                                                  0.000136
## educFour year college or university degree/Bachelor's degree (e.g., BS, BA, AB)                                          0.103930
## educHigh school graduate (Grade 12 with diploma or GED certificate)                                                      0.030437
## educHigh school incomplete (Grades 9-11 or Grade 12 with NO diploma)                                                     0.079269
## educLess than high school (Grades 1-8 or no formal schooling)                                                            0.013176
## educPostgraduate or professional degree, including master's, doctorate, medical or law degree (e.g., MA, MS, PhD, MD...) 0.107458
## educSome college, no degree (includes some community college)                                                            0.257834
## educSome postgraduate or professional schooling, no postgraduate degree (e.g. some graduate school)                      0.072411
## educTwo year associate degree from a college or university                                                               0.156001
## agerec35-39                                                                                                              0.253811
## agerec40-44                                                                                                              0.616329
## agerec45-49                                                                                                              0.010643
## agerec50-54                                                                                                              0.003870
## agerec55-59                                                                                                              0.000200
## agerec60-64                                                                                                              7.64e-07
## agerec65-69                                                                                                              7.70e-13
## agerec70-74                                                                                                              3.26e-08
## agerec75-79                                                                                                              2.01e-11
## agerec80-84                                                                                                              1.93e-10
## agerec85-89                                                                                                              4.55e-13
## agerecAge 24 or younger                                                                                                  0.543396
## agerecAge 25-29                                                                                                          0.463629
## agerecAge 90 or older                                                                                                    7.08e-08
## agerecDon't know/refused                                                                                                 0.009545
## maritalDivorced                                                                                                          0.133687
## maritalLiving with a partner                                                                                             0.347146
## maritalMarried                                                                                                           0.039560
## maritalNever been married                                                                                                0.010544
## maritalSeparated                                                                                                         0.118581
## maritalWidowed                                                                                                           0.014542
## attendA few times a year                                                                                                 0.163967
## attendMore than once a week                                                                                              0.160547
## attendNever                                                                                                              0.000255
## attendOnce a week                                                                                                        0.016552
## attendOnce or twice a month                                                                                              0.012887
## attendSeldom                                                                                                             0.137621
## religious_importanceNot at all important                                                                                 5.47e-08
## religious_importanceNot too important                                                                                    0.175246
## religious_importanceSomewhat important                                                                                   2.31e-12
## religious_importanceVery important                                                                                        < 2e-16
##                                                                                                                             
## (Intercept)                                                                                                              ***
## income$150,000 or more                                                                                                      
## income10 to under $20,000                                                                                                *  
## income100 to under $150,000                                                                                                 
## income20 to under $30,000                                                                                                ** 
## income30 to under $40,000                                                                                                   
## income40 to under $50,000                                                                                                   
## income50 to under $75,000                                                                                                   
## income75 to under $100,000                                                                                                  
## incomeLess than $10,000                                                                                                  ***
## educFour year college or university degree/Bachelor's degree (e.g., BS, BA, AB)                                             
## educHigh school graduate (Grade 12 with diploma or GED certificate)                                                      *  
## educHigh school incomplete (Grades 9-11 or Grade 12 with NO diploma)                                                     .  
## educLess than high school (Grades 1-8 or no formal schooling)                                                            *  
## educPostgraduate or professional degree, including master's, doctorate, medical or law degree (e.g., MA, MS, PhD, MD...)    
## educSome college, no degree (includes some community college)                                                               
## educSome postgraduate or professional schooling, no postgraduate degree (e.g. some graduate school)                      .  
## educTwo year associate degree from a college or university                                                                  
## agerec35-39                                                                                                                 
## agerec40-44                                                                                                                 
## agerec45-49                                                                                                              *  
## agerec50-54                                                                                                              ** 
## agerec55-59                                                                                                              ***
## agerec60-64                                                                                                              ***
## agerec65-69                                                                                                              ***
## agerec70-74                                                                                                              ***
## agerec75-79                                                                                                              ***
## agerec80-84                                                                                                              ***
## agerec85-89                                                                                                              ***
## agerecAge 24 or younger                                                                                                     
## agerecAge 25-29                                                                                                             
## agerecAge 90 or older                                                                                                    ***
## agerecDon't know/refused                                                                                                 ** 
## maritalDivorced                                                                                                             
## maritalLiving with a partner                                                                                                
## maritalMarried                                                                                                           *  
## maritalNever been married                                                                                                *  
## maritalSeparated                                                                                                            
## maritalWidowed                                                                                                           *  
## attendA few times a year                                                                                                    
## attendMore than once a week                                                                                                 
## attendNever                                                                                                              ***
## attendOnce a week                                                                                                        *  
## attendOnce or twice a month                                                                                              *  
## attendSeldom                                                                                                                
## religious_importanceNot at all important                                                                                 ***
## religious_importanceNot too important                                                                                       
## religious_importanceSomewhat important                                                                                   ***
## religious_importanceVery important                                                                                       ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 45368  on 34889  degrees of freedom
## Residual deviance: 39146  on 34841  degrees of freedom
## AIC: 39244
## 
## Number of Fisher Scoring iterations: 4

Key Findings:

• Significant Predictors: Variables such as income levels (especially the lowest income bracket), educational attainment, age categories, marital status, frequency of attending religious services, and the importance placed on religion are significant predictors in your model. These factors exhibit a range of influences, with some increasing and others decreasing the likelihood of changing religious affiliation.

  For example:

  • Lower income and older age groups significantly predict changes in religious beliefs.
  • Higher levels of religious importance significantly decrease the likelihood of changing religious affiliation.

Visualization and Further Analysis:

To better understand the model performance and to ensure the robustness of the findings, consider visualizing the following:

• Residual Plots: Check residuals to assess any systematic deviations that could indicate model mis-specifications.
• ROC Curve: Generate an ROC curve to evaluate the diagnostic ability of the model across various classification thresholds.
• Predicted vs. Actual: Plotting predicted probabilities against actual outcomes can help visualize the model’s accuracy in predicting changes in religious affiliation.

# Extract residuals
residuals_data <- data.frame(Residuals = residuals(glm_model, type = "deviance"))

# For logistic regression, residuals won't be perfectly normal, but let's visualize:

ggplot(residuals_data, aes(x = seq_along(Residuals), y = Residuals)) +
  geom_point(alpha = 0.5) +
  labs(x = "Index", y = "Residuals", title = "Plot of Residuals") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5))

p <- ggplot(residuals_data, aes(x = Residuals)) +
  geom_histogram(aes(y = ..density..), bins = 30, fill = "blue", alpha = 0.5) +
  geom_density(color = "red") +
  labs(x = "Residuals", y = "Density", title = "Histogram and Density Plot of Residuals") +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5))

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 600)

# Show the interactive plot in your default browser
interactive_plot

These observations can inform us about the model fit. While a degree of asymmetry in residuals can be expected in logistic regression, significant deviations from the expected distribution may suggest model mis-specifications, non-linearity in the data that the model hasn’t captured.

ROC curve

# Assuming you have a binary outcome
probabilities <- predict(glm_model, type = "response")
roc_curve <- roc(response = data_clean_2014$Religion_Change, predictor = probabilities)

## Setting levels: control = Changed, case = Same

## Setting direction: controls < cases

# Create a data frame for ggplot
roc_data <- data.frame(
  FPR = 1- roc_curve$specificities,
  TPR = roc_curve$sensitivities
)

# Plot ROC curve using ggplot2
 ggplot(roc_data, aes(x = FPR, y = TPR)) +
  geom_line(color = "darkorange") +
  geom_abline(linetype = "dashed") +
  labs(x = "False Positive Rate", y = "True Positive Rate", title = "ROC Curve") +
  theme_minimal() +
  geom_area(alpha = 0.2) +
  geom_segment(aes(x = 0, y = 0, xend = 1, yend = 1), color = "navy", linetype = "dashed") +
  annotate("text", x = 0.5, y = 0.25, label = paste("AUC =", round(auc(roc_curve), 2)), color = "red")

With an AUC of 0.73, the model has good predictive ability, significantly better than random guessing, which would have an AUC of 0.5. An AUC value in the range of 0.7 to 0.8 indicates that the model has a fair discrimination ability to differentiate between the two classes: those who have changed their religious affiliation and those who have not.

# Create a dataframe for ggplot
predictions_data <- data.frame(
  Actual = data_clean_2014$Religion_Change,
  Predicted_Probability = probabilities
)

# Plot Predicted Probabilities vs Actual Outcomes
p <- ggplot(predictions_data, aes(x = Predicted_Probability, fill = Actual)) +
  geom_histogram(position = "identity", alpha = 0.5, bins = 30) +
  labs(x = "Predicted Probability", y = "Count", title = "Predicted vs Actual Outcomes") +
  scale_fill_brewer(palette = "Set1", name = "Actual Outcome") +
  theme_minimal() +
  theme(legend.position = "bottom", plot.title = element_text(hjust = 0.5))

# Make it interactive
interactive_plot <- ggplotly(p , width = 1200 , height = 600)

# Show the interactive plot in your default browser
interactive_plot

The plot reveals that our model is more adept at predicting instances where individuals retain their childhood religious affiliation (‘Same’ class) than instances where they change their religious beliefs (‘Changed’ class). This discrepancy might stem from an imbalance in the dataset, where there are significantly more cases of one class compared to the other. As a result, the model may have learned the patterns associated with the ‘Same’ class more effectively because there were more examples to learn from, leading to a less robust understanding of the ‘Changed’ class characteristics. Addressing this imbalance could potentially improve the model’s ability to discern the subtler patterns that indicate a shift in religious affiliation.

Conclusion

In Chapter 4, I embarked on an analytical journey to model the dynamics of religious affiliation changes using a transition probability approach. Our exploration aimed to quantify how individuals transition between religious affiliations from childhood to adulthood and to simulate these changes over time within a societal context.

Key Findings:

• Transition Probabilities: successfully quantified the probabilities of transitioning from one religious affiliation to another. The transition matrix constructed from these probabilities revealed detailed insights into the patterns of religious switching. Notably, significant transitions were observed towards non-affiliation (‘Nothing in particular’), and there was an evident decline in traditional affiliations such as Protestantism and Catholicism.

• Simulation Insights: The simulation of religious composition over time highlighted potential trends and shifts in religious demographics. It demonstrated how the religious landscape could evolve, emphasizing the growing prevalence of religious non-affiliation alongside the decreasing proportions of established denominations.

• Steady State Analysis: Our analysis sought to determine if the distribution of religious affiliations would reach a steady state, where the proportions of each religion stabilize over time. The findings suggested that while some form of equilibrium could be anticipated, the religious landscape remains dynamically influenced by various social, cultural, and personal factors.

Methodological Insights:

• My methodology involved calculating transition probabilities and applying these to simulate the evolution of religious affiliations in a hypothetical population. This approach provided a framework for understanding the fluidity and changing nature of religious identities.

• The visualization techniques employed, including heatmaps of transition probabilities and time-series graphs of religious compositions, offered a clear and engaging representation of our findings, facilitating a deeper understanding of the underlying dynamics.

Challenges and Limitations:

• One of the primary challenges was the inherent data imbalance in religious affiliations, which influenced the model’s predictions and interpretations, particularly underrepresenting minority religious groups.

• Additionally, the model assumptions did not account for external factors such as immigration, birth, and death rates, which are crucial in real-world demographic studies. Future models would benefit from integrating these elements to provide more comprehensive insights.

Implications and Future Directions:

• The study’s findings are crucial for policymakers, religious leaders, educators, and sociologists as they provide a quantified insight into how religious affiliations are changing over generations. This knowledge can help in crafting strategies that address the needs and preferences of diverse religious groups.

• Future research could expand upon this foundational work by incorporating more complex demographic models and exploring the impact of socio-political changes on religious affiliations. Further, qualitative studies could complement this quantitative analysis to delve deeper into the reasons behind religious switching.

Refrences

To properly cite the 2014 and 2007 Pew Research Center datasets in APA format, you can structure your references like this:

References

For the 2014 Dataset:

Pew Research Center. (2015). America’s Changing Religious Landscape. Retrieved from https://www.pewresearch.org/religion/2015/05/12/americas-changing-religious-landscape/

For the 2007 Dataset:

Pew Research Center. (2008). U.S. Religious Landscape Survey: Religious Beliefs and Practices. Retrieved from https://www.pewresearch.org/religion/2008/06/01/u-s-religious-landscape-survey-religious-beliefs-and-practices/